System

Abstract

To provide a system.SOLUTION: A system includes: means for collecting plan and service information of its own company and other companies, and dialogue contents of excellent employees of its own company, and accumulating them in a database; means for preprocessing the accumulated data and training using a generative AI model; means for analyzing an inquiry from a user and generating an optimum answer using the generative AI model; and means for providing a terminal of the user with the generated answer.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including a description and related instruction sentence regarding the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2022-180282 A Summary of the Invention [Problem to be solved by the invention]

[0004] The challenge is to generate talk that differentiates the company from other companies when it releases a new plan, and to generate counter talk to compete when other companies release new plans. In addition, it is also a challenge to create new plans and service proposals that exploit the weaknesses of other companies, to use it as a conversation wall during training for new crew members, and to improve acquisition efficiency by transferring acquisition to robots in the future. [Means for solving the problem]

[0005] By using a method to have AI learn all plans and services of the company and other companies, as well as the talk of the company's excellent crew, when the company releases a new plan, it generates talk that differentiates it from other companies. It also generates counter talk when other companies release a new plan. By using these methods, it can be used to create new plans and services that exploit the weaknesses of other companies, and to bounce ideas off of new crew members during training. Furthermore, by shifting acquisition to robots in the future, it will be possible to improve acquisition efficiency. [Brief description of the drawings]

[0006] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Diagram 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. FIG. [Diagram 3] FIG. 11 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Diagram 5] FIG. 13 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 13 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 13 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 2 is a sequence diagram showing a process flow of the data processing system in the first embodiment of the first form example. [Figure 12]FIG. 11 is a sequence diagram showing a process flow of the data processing system in the application example 1 of the first embodiment. [Figure 13] FIG. 11 is a sequence diagram showing a process flow of the data processing system in the second embodiment of the second embodiment. [Figure 14] FIG. 13 is a sequence diagram showing a process flow of the data processing system in application example 2 of embodiment 2. [Figure 15] FIG. 13 is a sequence diagram showing a process flow of a data processing system in a third embodiment of the present invention. [Figure 16] FIG. 13 is a sequence diagram showing a process flow of the data processing system in the third application example of the third embodiment. [Figure 17] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in the first embodiment of the first form example when an emotion engine is combined. [Figure 18] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Form Example 1 when combined with an emotion engine. [Figure 19] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system according to the embodiment 2 of the second embodiment in the case where an emotion engine is combined. [Figure 20] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Form Example 2 when combined with an emotion engine. [Figure 21] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in the embodiment 3 of the form example 3 in the case where an emotion engine is combined. [Figure 22] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Form Example 3 when combined with an emotion engine. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0007] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0008] First, the terms used in the following description will be explained.

[0009] In the following embodiments, a signed processor (hereinafter simply referred to as a "processor") may be one arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be one type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), or a TPU (TENSOR PROCESSING UNIT (registered trademark)).

[0010] In the following embodiments, a signed RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by the processor.

[0011] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0012] In the following embodiments, a communication I / F (Interface) with a code is an interface including a communication processor and an antenna. The communication I / F controls communication between multiple computers. An example of a communication standard applied to the communication I / F is a wireless communication standard including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0013] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. In addition, in this specification, the same idea as "A and / or B" is also applied when three or more things are expressed by connecting them with "and / or."

[0014] [First embodiment]

[0015] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0016] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0017] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a wide area network (WAN) and / or a local area network (LAN).

[0018] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a

[0019] The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0020] The reception device 38 includes a touch panel 38A and a microphone 38B, and receives user input. The touch panel 38A detects contact with a pointer (e.g., a pen or a finger) to receive user input by the touch of the pointer. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0021] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (e.g., voice and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs voice according to instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a Complementary Metal-Oxide-Semiconductor (CMOS) image sensor or a Charge Coupled Device (CCD) image sensor.

[0022] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54.

[0023] FIG. 2 shows an example of main functions of the data processing device 12 and the smart device 14.

[0024] As shown in Fig. 2, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32. The specific process program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific process program 56 from the storage 32, and executes the read specific process program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific process program 56 executed on the RAM 30.

[0025] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0026] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores a reception output program 60. The reception output program 60 is used together with the specific processing program 56 by the data processing system 10. The processor 46 reads out the reception output program 60 from the storage 50, and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0027] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0028] "Example 1"

[0029] The present invention has a means for making AI learn all plans and services of the company and other companies, and the talk of the company's excellent crew. Specifically, information on plans and services of the company and other companies, and the contents of the talk of the company's excellent crew are stored in a database, and the AI ​​learns this information.

[0030] "Example 2"

[0031] Furthermore, when a company releases a new plan, it has the means to generate a differentiating message from other companies. Specifically, based on the information the AI ​​has learned, it generates a proposal message for a new plan that is differentiated from other companies' plans and services. For example, if another company's plan is "unlimited calling charges for 1,000 yen per month," the AI ​​generates a differentiating message that says "unlimited calling charges and unlimited data communication for 500 yen per month."

[0032] "Example 3"

[0033] In addition, it has a means to generate counter talk when other companies release new plans. Specifically, when another company releases a new plan, the AI ​​learns information about that plan and generates counter talk to counter it. For example, if another company releases a plan that offers "unlimited calls and unlimited data for 500 yen per month," the AI ​​generates the counter talk "unlimited calls and unlimited data for 500 yen per month, plus 24 / 7 support."

[0034] "Example 4"

[0035] In addition, it has the means to create new plans and service proposals that exploit the weaknesses of other companies' plans and services, and to use them as a means of bouncing ideas off new crew members during training. Specifically, based on the information that AI has learned, it creates new plans and service proposals that exploit the weaknesses of other companies' plans and services. In addition, during training of new crew members, the talks generated by AI are used as material for bouncing ideas off new crew members.

[0036] "Example 5"

[0037] Finally, in the future, we will have the means to improve the efficiency of customer acquisition by transferring the acquisition to robots. Specifically, the robot will use the chat generated by AI to improve the efficiency of customer acquisition. For example, the robot will automatically conduct sales activities to customers based on the chat generated by AI.

[0038] The process flow of each embodiment will be described below.

[0039] "Example 1"

[0040] Step 1: Accumulate a database of all plans and services from your company and other companies, as well as information on the talks of your company's top crew members.

[0041] Step 2: The AI ​​learns the information from the database.

[0042] "Example 2"

[0043] Step 1: Based on the information the AI ​​has learned, it generates a proposal speech for a new plan that is differentiated from other companies' plans and services.

[0044] Step 2: Provide the generated chat to the sales team and use it to propose new plans.

[0045] "Example 3"

[0046] Step 1: When another company releases a new plan, the AI ​​learns the information about that plan.

[0047] Step 2: Based on the information that the AI ​​has learned, it generates a counter talk to counter the new plan. Step 3: The generated counter talk is provided to the sales team and used in sales activities to counter the new plan of other companies.

[0048] "Example 4"

[0049] Step 1: Based on the information the AI ​​has learned, it creates new plan and service proposals that exploit the weaknesses of other companies' plans and services.

[0050] Step 2: When training new crew members, use the AI-generated talk as a training material.

[0051] "Example 5"

[0052] Step 1: The robot uses AI-generated chat to improve customer acquisition efficiency.

[0053] Step 2: Based on the conversation generated by the AI, the robot automatically conducts sales activities with the customer.

[0054] Example 1

[0055] Next, a description will be given of Example 1 of the embodiment 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal".

[0056] With the previous system, it was difficult to effectively utilize plan and service information from the company and other companies, as well as the dialogue content of excellent employees, making it difficult to provide quick and appropriate answers to user inquiries. In addition, efficient methods were required for creating new plan and service proposals and training new employees. Furthermore, there was a lack of automated means to improve the efficiency of customer acquisition.

[0057] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0058] In this invention, the server includes a means for collecting plan and service information of the company and other companies, and the contents of dialogue of the company's excellent employees, and storing them in a database, a means for preprocessing the stored data and learning them using a generative AI model, and a means for analyzing inquiries from users and generating optimal answers using the generative AI model. This makes it possible to provide prompt and appropriate answers to inquiries from users. It also provides an efficient method for creating new plan and service proposals and training new employees. Furthermore, it is possible to realize an automation means for improving the efficiency of customer acquisition.

[0059] "Plan and service information" is detailed data on the various services and pricing plans offered by companies.

[0060] "Conversations between excellent employees" are records of actual conversations conducted by employees who were highly evaluated for their customer service.

[0061] A "database" is an information accumulation system that systematically stores collected data and enables efficient search and use.

[0062] "Preprocessing" refers to processes such as data cleaning, normalization, and tokenization to prepare collected data in a format suitable for an AI model.

[0063] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate answers to user inquiries.

[0064] A "user query" is a question or information request made by a user to the system.

[0065] "Answer generation" is the process of creating an answer based on appropriate information to a user's inquiry.

[0066] A "terminal" is a device through which a user accesses and queries the system.

[0067] "Dialogue simulation" is a training method that allows new employees to gain simulated experience interacting with real customers.

[0068] "Customer acquisition automation" is a technology that automates the process of efficiently acquiring customers.

[0069] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0070] The present invention is a system for providing prompt and appropriate responses to inquiries from users. A specific embodiment of this system will be described below.

[0071] 1. Program Generation

[0072] The server generates a program that collects the plan and service information of the company and other companies provided by the user, as well as the conversations of the company's outstanding employees, and stores them in a database. This program is developed using programming languages ​​such as Python and Java (registered trademark).

[0073] 2. Data collection and storage

[0074] The server stores the collected data in a database. A relational database such as MySQL (registered trademark) or PostgreSQL (registered trademark) is used as the database. Specifically, the following data is collected:

[0075] Detailed information about your plans and services (e.g., pricing, offerings, terms and conditions)

[0076] Detailed information about other companies' plans and services (e.g., fees, services provided, contract terms)

[0077] Conversations by your company's outstanding employees (e.g. customer service scripts, success stories)

[0078] 3. Data Preprocessing

[0079] The server preprocesses the data stored in the database, specifically cleaning the data (filling in missing values ​​and removing outliers), normalizing it, tokenizing it, etc. This prepares the data in a format suitable for the generative AI model.

[0080] 4. Training the model

[0081] The server trains a generative AI model using the preprocessed data. The AI ​​model used is a natural language processing model such as GPT-3 (registered trademark) or BERT. The learning process uses machine learning frameworks such as TENSORFLOW (registered trademark) or PyTorch (registered trademark).

[0082] 5. Parsing the query and generating an answer

[0083] When a user makes an inquiry through a terminal, the server analyzes the content of the inquiry. It uses natural language processing technology to understand the user's intent and extract the necessary information. The server uses a generative AI model based on the analysis results to generate a response. Specifically, it combines the optimal plan and service information and dialogue content for the user's inquiry to create a response in a natural dialogue format.

[0084] 6. Providing a response

[0085] The server sends the generated answer to the user's terminal, through which the user can check the answer and make additional inquiries if necessary.

[0086] Examples

[0087] For example, if a user asks "What is the difference between my company's plan A and another company's plan B?" the server would process it as follows:

[0088] 1. Analyze the user's query and retrieve information for Plan A and Plan B from the database.

[0089] 2. Compare the information obtained and clarify the differences.

[0090] 3. Generate answers in a natural, conversational format based on the conversations of your top employees.

[0091] Examples of prompt statements

[0092] Examples of prompts to be fed to a generative AI model include:

[0093] User: "What is the difference between my Plan A and the other company's Plan B?"

[0094] AI model: "Our company's Plan A costs 5,000 yen per month and has unlimited data. The other company's Plan B costs 4,500 yen per month and has a data limit of 10GB. In addition, Plan A includes free calls, but Plan B does not."

[0095] In this way, the server realizes a system that provides appropriate answers to user inquiries.

[0096] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0097] Step 1: Data collection

[0098] The server collects plan and service information from the company and other companies, as well as the content of conversations between the company's outstanding employees. Specifically, it uses APIs and scraping tools to obtain the necessary data from websites and internal systems. The inputs include various API endpoints and webpage URLs, and the output is the acquired plan and service information and conversation content data.

[0099] Step 2: Data accumulation

[0100] The server stores the collected data in a database. MySQL (registered trademark) or PostgreSQL (registered trademark) is used as the database. Specifically, it executes SQL queries to insert the acquired data into appropriate tables. The inputs include the collected plan and service information and conversation content data, and the output is the data stored in the database.

[0101] Step 3: Data Preprocessing

[0102] The server preprocesses the data stored in the database, specifically cleaning the data (filling in missing values ​​and removing outliers), normalizing it, tokenizing it, etc. The input is the raw data stored in the database, and the output is the preprocessed, clean data.

[0103] Step 4: Model training

[0104] The server trains a generative AI model using the preprocessed data. The AI ​​models used are GPT-3 (registered trademark) and BERT, and training is performed using TensorFlow (registered trademark) and PyTorch (registered trademark). Specifically, the data is divided into a training set and a test set, and the accuracy of the model is evaluated. The input is the preprocessed data, and the output is a trained AI model.

[0105] Step 5: Query Analysis

[0106] When a user makes a query through a terminal, the server analyzes the query. It uses natural language processing technology to understand the user's intent and extract the necessary information. The input is the query text from the user, and the output is the analyzed query.

[0107] Step 6: Answer Generation

[0108] The server generates an answer using a generative AI model based on the analysis results. Specifically, it creates an answer in a natural dialogue format by combining the optimal plan and service information and dialogue content for the user's inquiry. The input is the analyzed inquiry content and the trained AI model, and the output is the generated answer text.

[0109] Step 7: Provide your answer

[0110] The server sends the generated answer to the user's terminal, where the user can check the answer and make additional inquiries if necessary. The input is the generated answer text, and the output is the answer displayed on the user's terminal.

[0111] (Application example 1)

[0112] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0113] With the conventional system, it was difficult to effectively utilize plan and service information from our company and other companies to make optimal proposals to users. In addition, there was a lack of a way to provide optimal advice to users through chat using AI that learned the conversation content of excellent crew members. This resulted in a lack of improvement in the user's purchasing experience, and there were concerns about a decline in competitiveness.

[0114] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0115] In this invention, the server includes a means for making an AI learn all plans and services of the company and other companies, and the talk of the company's top crew, a means for generating talk that differentiates the company from other companies when the company releases a new plan, a means for generating counter talk when other companies release a new plan, a means for proposing the best plan or service for a product that the user is considering purchasing, a means for providing the user with the best option based on the company's and other companies' plan and service information, and a means for the AI ​​that has learned the talk content of the company's top crew to provide the best advice in a chat with the user. This makes it possible to propose the best plan or service to the user and improve the purchasing experience.

[0116] "Plan / Service" refers to the content of the products and services offered by the company and their terms of use.

[0117] "Excellent crew" refers to employees within a company who have particularly outstanding performance and skills.

[0118] "Talk" refers to the way of speaking and the content of explanations used in customer interactions and sales activities.

[0119] "AI" is an abbreviation for artificial intelligence and refers to systems that mimic human intelligence using technologies such as machine learning and natural language processing.

[0120] "Differentiation talk" refers to the way of speaking and the content of explanations that emphasize the superiority of a company's plans and services compared to those of other companies.

[0121] "Countertalk" refers to the way of speaking or the explanation given when another company announces a new plan or service.

[0122] "User" refers to a customer or consumer who uses the company's plans or services.

[0123] "Chat" refers to a means of communicating in real time using text messages.

[0124] "Buying experience" refers to the overall experience and satisfaction a user feels when purchasing a product or service.

[0125] "Choices" refers to the selection of plans or services available to a User.

[0126] The system for implementing this invention has the following configuration. The server accumulates information on plans and services from the company and other companies, as well as the contents of the conversations of the company's top crew members in a database, and has the AI ​​learn this information. When a user tries to purchase a product using their smartphone, the server receives the user's input and suggests the most suitable plan and service.

[0127] Hardware and Software Used

[0128] Hardware: Smartphones, servers

[0129] Software: Python, OpenAI® API

[0130] Data processing and calculation

[0131] The server processes the data in the following steps:

[0132] 1. Building the database:

[0133] The server stores information on plans and services from the company and other companies, as well as the contents of the conversations of the company's outstanding crew members in a database, allowing the AI ​​to learn from this information.

[0134] 2. Receiving user input:

[0135] The user inputs information about the product he or she is considering purchasing using a smartphone. For example, the user might input, "I want to buy a new smartphone."

[0136] 3. Generate prompt:

[0137] The server generates prompts based on user input, including information about plans and services from the company and other companies, and the content of talks by outstanding crew members.

[0138] 4. AI-generated responses:

[0139] The server sends the generated prompts to the OpenAI® API and receives an AI-generated response, which includes recommendations for the best plan or service for the user.

[0140] 5. Display response to user:

[0141] The server then displays the response on the user's smartphone, allowing the user to select the most suitable plan or service.

[0142] Examples

[0143] If the user types "I want to buy a new smartphone", the server generates a prompt like this:

[0144] Example prompt:

[0145] What people are considering buying: A new smartphone

[0146] Please suggest the best plan and service for you.

[0147] Company Plans & Services: Plan A, Plan B, Plan C

[0148] Other companies' plans and services: Plan X, Plan Y, Plan Z

[0149] Topic of the excellent crew: This plan is very cost-effective. This service is especially recommended for beginners. This product is currently on sale and is a great deal.

[0150] Based on these prompts, the AI ​​can generate responses suggesting the best plan or service for the user, for example, "Plan A is currently on sale and is a good deal. Plan B is recommended for beginners."

[0151] This way, users can choose the best plan or service for them, improving their purchasing experience.

[0152] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0153] Step 1:

[0154] The server stores the plan and service information of the company and other companies, as well as the contents of the conversations of the company's outstanding crew members in a database. This allows the AI ​​to learn from this information. The input is the plan and service information and the contents of the conversations, and the output is a database in which this information is stored. In concrete terms, the server uses a database management system (DBMS) to organize the information and save it in an appropriate format.

[0155] Step 2:

[0156] A user uses a smartphone to input information about a product he or she is considering purchasing. For example, the input may be, "I want to buy a new smartphone." The input includes information about the product the user is considering, and the output is that information is sent to a server. Specifically, the user inputs text through a smartphone application, and the data is sent to the server.

[0157] Step 3:

[0158] The server generates a prompt based on the user's input. The prompt includes the company's and other companies' plan and service information, and the talk content of outstanding crew members. The input is the user's input information, as well as the plan and service information and talk content stored in the database, and the generated prompt is obtained as the output. In terms of specific operation, the prompt is constructed using a Python script.

[0159] Step 4:

[0160] The server sends the generated prompt to the OpenAI (registered trademark) API and obtains a response from the AI. The generated prompt is the input, and the AI's response is the output. The specific operation is to call the OpenAI (registered trademark) API, send a prompt, and receive a response.

[0161] Step 5:

[0162] The server displays the obtained response on the user's smartphone. This allows the user to select the most suitable plan and service. The input is the response from the AI, and the output is the information displayed on the user's smartphone. The specific operation is to update the UI (user interface) to display the response through the smartphone application.

[0163] In this way, users can select the most suitable plan or service, improving their purchasing experience.

[0164] Example 2

[0165] Next, a second embodiment of the second embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0166] In the conventional system, there was a problem that it took a lot of time and effort to generate counter talks to differentiate plans and services between one's own and those of other companies. In addition, it was difficult to quickly generate counter talks when other companies announced new plans, making it difficult to maintain competitiveness. Furthermore, there was an issue of a lack of practice in effective talks in training new employees.

[0167] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0168] In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and talk of the company's outstanding employees, a means for collecting plan and service information of other companies, a means for preprocessing the collected data, a means for inputting the preprocessed data into a generating AI model, a means for generating a differentiation talk from other companies using the generating AI model, a means for outputting the generated differentiation talk, and a means for generating a counter talk when another company issues a new plan. This makes it possible to generate differentiation talk and counter talk quickly and efficiently.

[0169] "All plans and services of our company and competitors" refers to all pricing plans and service contents offered by our company and competitors.

[0170] "Excellent employee talk" refers to the effective speaking techniques and sales pitches that a company's top employees use when dealing with customers and in sales activities.

[0171] "Artificial intelligence" refers to computer systems that mimic human intelligence using techniques such as machine learning and natural language processing.

[0172] "Means of collecting information on the plans and services of other companies" refers to methods of automatically obtaining competitors' pricing plans and service details using web scraping tools, APIs, etc.

[0173] "Means for preprocessing collected data" refers to methods for analyzing acquired data and processing it to delete unnecessary information or standardize the format.

[0174] "Means of inputting preprocessed data into a generative AI model" refers to a method of providing preprocessed data in an appropriate format to a generative AI model.

[0175] A "generative AI model" refers to an artificial intelligence model that generates new text or information based on input data.

[0176] "Means for generating differentiating talk" refers to a method of using a generative AI model to generate a speaking style or sales pitch to propose a new company's plan that is differentiated from the plans and services of other companies.

[0177] "Means for outputting the generated differentiated talk" refers to a method for providing a user with the differentiated talk created by the generative AI model.

[0178] "Means of generating countertalk" refers to the way in which you generate speech or sales pitches to counter an announcement of a new plan by another company.

[0179] The present invention is a system in which a server, a terminal, and a user work together, and is implemented as follows.

[0180] System Configuration

[0181] The server has the following main functions:

[0182] 1. Data collection function:

[0183] The server collects plan and service information from other companies. This information is obtained through web scraping tools (e.g. BeautifulSoup, Scrapy) and APIs.

[0184] 2. Data pre-processing function:

[0185] The server preprocesses the collected data, specifically cleaning the text data (removing unnecessary characters and tags) and normalizing it (unifying uppercase and lowercase letters, converting special characters).

[0186] 3. Input functions for generative AI model:

[0187] The server inputs the preprocessed data into a generative AI model (e.g., GPT-4 (registered trademark)) by providing instructions to the model using prompt sentences.

[0188] 4. Differentiated Talk Generation Function:

[0189] The generative AI model generates talk to differentiate a company's new plan based on the input information about other companies' plans.

[0190] 5. Result output function:

[0191] The server outputs the generated differentiated token and transmits it to the terminal.

[0192] Hardware and Software Used

[0193] Web scraping tools: BeautifulSoup, Scrapy

[0194] Generative AI model: GPT-4 (registered trademark)

[0195] Data preprocessing tools: Python standard library and NLP libraries (e.g. NLTK, spaCy)

[0196] Examples

[0197] Example 1:

[0198] Other companies' plan information: "Unlimited calls for 1,000 yen per month"

[0199] Generated differentiated talk: "Unlimited calls and unlimited data for just 500 yen per month."

[0200] Example prompt:

[0201] "The other company's plan is 1,000 yen per month with unlimited calls. Please generate a chat to differentiate our new plan from the others."

[0202] How it works

[0203] 1. Data collection:

[0204] The server accesses the specified URL and retrieves the HTML content. It then uses BeautifulSoup to parse the HTML and extract the necessary information (plan name, price, service details, etc.).

[0205] 2. Data preprocessing:

[0206] The server removes HTML tags from the retrieved text data, removes extra spaces and line breaks, converts all characters to lower case, and converts special characters to their standard form.

[0207] 3. Input to the generative AI model:

[0208] The server inputs the preprocessed data into a generative AI model, providing instructions to the model using prompts.

[0209] 4. Generating Differentiated Talk:

[0210] The generative AI model analyzes information about other companies' plans and generates talk to propose new, competitive plans.

[0211] 5. Result output:

[0212] The server saves the chat received from the generative AI model in text format and sends it to the user's device. The user can review the chat and use it to propose new plans for their company.

[0213] In this way, the server utilizes generative AI models to provide a system that efficiently generates differentiated talks from other companies.

[0214] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0215] Step 1:

[0216] The server collects plan and service information from other companies.

[0217] Input: Third-party website URL or API endpoint.

[0218] Specific operation: The server accesses the specified URL and retrieves the HTML content. Then, it parses the HTML using BeautifulSoup or Scrapy and extracts the necessary information (plan name, price, service content, etc.).

[0219] Output: A list of extracted plans and services from other companies.

[0220] Step 2:

[0221] The server pre-processes the collected data.

[0222] Input: Third-party plan / service information collected in step 1.

[0223] What happens: The server removes HTML tags from the retrieved text, removes extra spaces and line breaks, converts all characters to lower case, and converts special characters to their standard form.

[0224] Output: Preprocessed and clean text data.

[0225] Step 3:

[0226] The server inputs the preprocessed data into a generative AI model.

[0227] Input: The clean text data preprocessed in step 2.

[0228] Specific operation: The server inputs the preprocessed data into a generative AI model (e.g., GPT-4 (registered trademark)). At this time, it gives instructions to the model using a prompt sentence. For example, it uses a prompt sentence such as, "Competitor's plans are 1,000 yen per month with unlimited calls. In contrast, please generate a chat to differentiate our new plan."

[0229] Output: The data and prompts fed into the generative AI model.

[0230] Step 4:

[0231] The generative AI model generates talk to differentiate a company's new plan based on the input information about other companies' plans.

[0232] Input: The data and prompts fed into the generative AI model in step 3.

[0233] Specific operation: The generative AI model analyzes the plan information of other companies and generates a message to propose a new competitive plan. For example, it generates a message saying, "Unlimited calls and unlimited data for 500 yen per month."

[0234] Output: The generated differentiation talk.

[0235] Step 5:

[0236] The server outputs the generated differentiated token and transmits it to the terminal.

[0237] Input: Differentiated tokens generated in step 4.

[0238] Specific operation: The server saves the chat received from the generative AI model in text format and sends it to the user's device. The user can review the chat and use it to propose new plans for their company.

[0239] Output: Differentiated token sent to user's device.

[0240] (Application example 2)

[0241] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0242] In conventional systems, it is difficult to generate proposal talks for new plans to differentiate from competitors, and there is a lack of means for operators of online shopping sites to effectively differentiate from competitors. In addition, there is a lack of a way to easily share the generated proposal talks, which makes it difficult to respond quickly.

[0243] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0244] In this invention, the server includes a means for making an AI learn all plans and services of the company and other companies, and talks of the company's excellent crew, a means for generating talks that differentiate the company from other companies when the company issues a new plan, a means for generating counter talk when another company issues a new plan, a means for receiving a competitor's plan as an input and generating a proposal talk for the differentiated new plan, and a means for displaying the generated proposal talk to users and making it easy to share. This makes it possible to quickly and effectively generate and share proposal talks for new plans to differentiate the company from competitors.

[0245] "Plan" refers to the content and pricing of services or products offered by a company.

[0246] "Service" refers to any intangible value or benefit provided to a customer.

[0247] "Talk" refers to explanations or proposals given to customers in sales and marketing.

[0248] "AI" is an abbreviation for artificial intelligence, and refers to the technology that allows machines to imitate human intelligence to learn and make inferences.

[0249] "Differentiation talk" refers to explanations and proposals that emphasize the advantages of your company compared to your competitors' plans and services.

[0250] "Countertalk" refers to explanations or proposals made to counter a new plan announced by a competitor.

[0251] "Competitors" refers to other companies in the same market that offer similar plans or services.

[0252] A "proposal talk" refers to an explanation or suggestion given when proposing a new plan or service to a customer.

[0253] "User" refers to individuals or companies that use this system.

[0254] "Sharing" refers to sharing generated information or data with other users.

[0255] A system for implementing the present invention includes a server, a user terminal, and a generative AI model. The server receives input from the user terminal and generates differentiated talk and counter talk using the generative AI model. Specific embodiments are described below.

[0256] System Configuration

[0257] 1. Server

[0258] The server includes a database, an AI model, and an API. The database stores the plans and services of the company and other companies, and the talks of outstanding crew members. The AI ​​model is used to learn from these data and generate differentiated talks and counter talks. The API provides an interface for communication with user terminals.

[0259] 2. User terminal

[0260] The user terminal is a device such as a smartphone or a PC that is used by the user to input competitors' plans and check the generated tokens. The user terminal communicates with the server to send and receive the necessary data.

[0261] 3. Generative AI Models

[0262] Generative AI models use advanced natural language processing models such as OpenAI's GPT-3, which generate differentiation and countertalk based on a prompt.

[0263] Processing flow

[0264] 1. Data Entry

[0265] The user inputs a competitor's plan using the user terminal. For example, the user inputs a plan that offers "unlimited calls for 1,000 yen per month."

[0266] 2. Prompt generation

[0267] The server generates a prompt based on the input plan. An example of a prompt is "If the plan of another company is '1000 yen per month with unlimited calls,' please generate a proposal for a new differentiated plan."

[0268] 3. Creating a token

[0269] The server sends a prompt to the generative AI model to generate differentiated conversations. For example, the conversations generated might be something like, "Unlimited calls and unlimited data for 500 yen per month."

[0270] 4. View and share your results

[0271] The generated token is sent to the user's terminal and displayed to the user, who can then easily share the token.

[0272] Hardware and Software Used

[0273] Hardware: Servers and user devices (smartphones, PCs, etc.) that can connect to the Internet

[0274] Software: Python, OpenAI® API, database management system (e.g. MySQL®)

[0275] Examples

[0276] If the user enters a competitor's plan, "Unlimited calls for 1000 yen per month," the server generates the following prompt:

[0277] If another company's plan offers unlimited calls for 1,000 yen per month, create a pitch for a new, differentiated plan.

[0278] Based on this prompt, the generative AI model generates a differentiated conversation that says, "Unlimited calls and unlimited data for 500 yen per month," and displays it on the user's device.

[0279] In this way, users can quickly generate and share effective pitch talks that differentiate them from their competitors.

[0280] The flow of the specific process in the application example 2 will be described with reference to FIG.

[0281] Step 1:

[0282] User enters competitor's plan.

[0283] Input: Competitor's plan (e.g. "Unlimited calls for 1000 yen per month")

[0284] Output: The input plan data

[0285] Specific operation: The user uses a user device such as a smartphone or a PC to input the competitor's plan into an input form. The input data is then sent to the server.

[0286] Step 2:

[0287] The server receives the entered plan data and generates a prompt sentence.

[0288] Input: Plan data entered by the user

[0289] Output: Generated prompt text (e.g. "If the other company's plan is '1000 yen per month with unlimited calls,' please generate a proposal for a new differentiated plan.")

[0290] Specific operation: The server generates a prompt sentence to be sent to the generative AI model based on the received plan data. The prompt sentence includes instructions for generating a differentiation speech based on the competitor's plan.

[0291] Step 3:

[0292] The server sends prompts to the generative AI model to generate differentiated speech.

[0293] Input: Generated prompt text

[0294] Output: Generated differentiated talk (e.g. "Unlimited calls and unlimited data for 500 yen per month")

[0295] Specific operation: The server sends the generated prompt sentence to the generative AI model through the OpenAI (registered trademark) API. The generative AI model generates differentiated speech based on the prompt sentence and returns the result to the server.

[0296] Step 4:

[0297] The server transmits the generated differentiated token to the user terminal.

[0298] Input: Generated differentiation token

[0299] Output: Differentiated tokens displayed on the user's device

[0300] Specific operation: The server sends the differentiated token received from the generative AI model to the user terminal, which displays the token to the user.

[0301] Step 5:

[0302] Users can review the generated differentiated talks and share them if desired.

[0303] Input: Differentiated tokens displayed on the user's device

[0304] Output: Shared differentiation talk (e.g. share via email or social media)

[0305] Specific operation: The user checks the differentiated talk displayed on the user terminal and shares it with other users or related parties as necessary. Sharing is done through means such as email or SNS.

[0306] In this way, users can quickly generate and share effective pitch talks that differentiate them from their competitors.

[0307] Example 3

[0308] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0309] In the conventional system, it was difficult to generate counter talk quickly and effectively when other companies announced new plans. In addition, there was a lack of means to generate and utilize effective talks when training new crew members and improving the efficiency of customer acquisition. This led to concerns about a decline in competitiveness and a decline in customer satisfaction.

[0310] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0311] In this invention, the server includes a means for collecting information on new plans of other companies, a means for making an artificial intelligence learn the collected information, and a means for storing the generated counter talk in a database. This makes it possible to generate counter talk that quickly and effectively counters new plans of other companies. In addition, by including a means for using the talk generated during training of new crew members as material for bouncing ideas off each other, and a means for a robot to use the talk generated by the artificial intelligence to improve the efficiency of customer acquisition, it is possible to improve the training effect of new crew members and to improve the efficiency of customer acquisition.

[0312] "Artificial intelligence" is a computer system that has the ability to learn from data, recognize patterns, and generate new information.

[0313] "Countertalk" is persuasive dialogue generated to counter other companies' plans or services.

[0314] A "database" is a system for efficiently storing, managing, and retrieving information.

[0315] "New crew" refers to newly hired employees or staff.

[0316] "Beat the Wall" is a simulation technique used as an interlocutor during dialogue or training.

[0317] A "robot" is a mechanical device that acts automatically according to programmed instructions to perform specific tasks.

[0318] "Customer acquisition" is the activity of attracting new customers and encouraging them to sign contracts or make purchases.

[0319] A "plan" is a contract that specifies the terms and conditions for providing a specific service or product.

[0320] "Service Proposal" means a proposal outlining the specific content and conditions of the services to be provided.

[0321] "Collection" is the act of gathering necessary information or data.

[0322] "Learning" is the process of analyzing data and acquiring patterns and knowledge.

[0323] "Generation" is the act of producing new information or data.

[0324] This invention is a system that generates counter talks quickly and effectively when other companies announce new plans. It also contributes to training new crew members and improving the efficiency of customer acquisition.

[0325] Hardware and Software Used

[0326] Hardware: Servers, terminals, robots

[0327] Software: Generative AI models (e.g., GPT-4®), database management systems (e.g., MySQL®), natural language processing tools (e.g., spaCy)

[0328] Specific implementation of the system

[0329] 1. Collect information on new plans from other companies

[0330] The server uses a web scraping tool (e.g. BeautifulSoup) to collect information on new plans from other companies' websites. Specifically, the server periodically checks other companies' websites, and if a new plan is announced, the information is stored in the database.

[0331] Specific behavior: The server scrapes the page "https: / / example.com / new-plans", extracts the new plan information, and saves it to the database.

[0332] 2. Learning Information

[0333] The server inputs the new plan information collected by the server into a generative AI model (e.g., GPT-4 (registered trademark)) and trains it. The server obtains new plan information from the database and feeds it to the generative AI model.

[0334] Specific operation: The server retrieves plan information for "unlimited calls and data for 500 yen per month" from the database and inputs it into the generative AI model.

[0335] 3. Generating countertalk

[0336] The server uses generative AI models to generate counter talk to counter other companies' plans, which are then stored in a database for future use.

[0337] Specific operation: The server inputs a prompt to the generative AI model: "Another company has announced a plan that offers unlimited calls and unlimited data for 500 yen per month. Please generate countertalk to this plan." and stores the generated countertalk in a database.

[0338] 4. Creating new plans and service proposals

[0339] The server uses the generative AI model to create new plans and service proposals that exploit the weaknesses of other companies' plans. The generated new plans and service proposals are also stored in the database.

[0340] Specific operation: The server inputs a prompt to the generative AI model saying, "Other companies' plans have a weakness in that their support systems are insufficient. Please propose a new plan that addresses this weakness." and saves the new plan generated in the database.

[0341] 5. Support for training new crew members

[0342] When training new crew members, the terminals will use the AI-generated chats as material for them to bounce ideas off of. New crew members will communicate with the AI ​​through the terminals and receive practical training.

[0343] What it does: A new crew member uses a terminal to type the prompt, "Please generate a talk that explains the benefits of our new plan," and then practices the generated talk.

[0344] 6. Improved customer acquisition efficiency

[0345] The robot uses the AI-generated talk to automatically conduct sales activities with customers. The robot makes a presentation to the customer based on the generated talk, promoting contracts.

[0346] Specific operation: The robot makes sales calls to customers using the phrase, "For 500 yen per month, you get unlimited calls, unlimited data, and 24 / 7 support."

[0347] Examples and prompts

[0348] Example 1: Generating countertalk

[0349] Other companies' plans: "Unlimited calls and data for 500 yen per month"

[0350] Generated countertalk: "For just ¥500 a month, you get unlimited calls, unlimited data, and 24 / 7 support."

[0351] Examples of prompt statements

[0352] Another company has announced a plan that offers unlimited calls and data for 500 yen per month. Please generate a counter-talk to this plan.

[0353] Example 2: Creating a new plan / service proposal

[0354] Weakness of other companies' plans: Insufficient support system

[0355] The new plan created: "Unlimited calls and data for 600 yen per month, plus 24 / 7 support and chat support via a dedicated app."

[0356] Examples of prompt statements

[0357] The weak point of other companies' plans is the insufficient support system. Please propose a new plan that takes advantage of this weak point.

[0358] Example 3: Support for new crew training

[0359] Generated chat: "Dear customer, our new plan includes 24 / 7 support, so you can use our services with peace of mind."

[0360] Examples of prompt statements

[0361] Generate onboarding talks for new crew members explaining the benefits of our new plan.

[0362] In this way, the system performs specific processing at each step, generates counter talks and new plan / service proposals to compete with new plans from other companies, and helps improve the efficiency of training new crew members and customer acquisition. The flow of the specific processing in the third embodiment will be described with reference to FIG.

[0363] Step 1:

[0364] Collecting information on new plans from other companies

[0365] The server uses a web scraping tool (e.g. BeautifulSoup) to collect information about new plans from other companies' websites.

[0366] Input: Third-party website URL

[0367] Specific action: The server scrapes the page "https: / / example.com / new-plans" and extracts information about new plans.

[0368] Data processing: Convert the extracted information into structured data (e.g., JSON format).

[0369] Output: Structured new plan information

[0370] Step 2:

[0371] Learning Information

[0372] The information about new plans collected by the server is input into a generative AI model (e.g., GPT-4 (registered trademark)) and trained.

[0373] Input: Structured new plan information

[0374] Specific operation: The server retrieves plan information for "unlimited calls and data for 500 yen per month" from the database and inputs it into the generative AI model.

[0375] Data computation: Generative AI models analyze new plan information and learn features and benefits.

[0376] Output: A trained generative AI model

[0377] Step 3:

[0378] Generating counter talk

[0379] The server uses generative AI models to generate counter talk to rival plans.

[0380] Input: trained generative AI model, prompt

[0381] Specific operation: The server inputs a prompt to the generated AI model: "Another company has announced a plan that offers unlimited calls and unlimited data for 500 yen per month. Please generate a countertalk to this plan."

[0382] Data computation: A generative AI model analyzes prompts and generates countertalk.

[0383] Output: The generated countertalk

[0384] Step 4:

[0385] Saving Countertalk

[0386] The server stores the generated countertalk in a database.

[0387] Input: Generated countertalk

[0388] Specific behavior: The server stores the generated countertalk in a database.

[0389] Data processing: Converting countertalk into an appropriate format (e.g., text format).

[0390] Output: Countertalk stored in the database

[0391] Step 5:

[0392] Creating new plans and service proposals

[0393] The server uses the generative AI model to create new plans and service proposals that exploit the weaknesses of other companies' plans.

[0394] Input: trained generative AI model, prompt

[0395] Specific operation: The server inputs a prompt to the generative AI model: "Other companies' plans have a weakness in that their support systems are insufficient. Please propose a new plan that addresses this weakness."

[0396] Data calculation: A generative AI model analyzes the prompt text and generates new plans and service proposals.

[0397] Output: Generated new plans and service proposals

[0398] Step 6:

[0399] Support for training new crew members

[0400] When training new crew members, the device will use the AI-generated talk as material for training.

[0401] Input: Generated chat, prompt

[0402] Specific Action: A new crew member uses a terminal to type the prompt, "Please generate a talk that explains the benefits of our new plan."

[0403] Data computation: The generative AI model analyzes the prompt and generates the talk.

[0404] Output: The generated token

[0405] Step 7:

[0406] Improved customer acquisition efficiency

[0407] The robot will use AI-generated chatter to automatically conduct sales activities with customers.

[0408] Input: The generated token

[0409] Specific operation: The robot makes sales calls to customers using the phrase, "For 500 yen per month, you get unlimited calls, unlimited data, and 24 / 7 support."

[0410] Data calculation: The robot analyzes the talk and presents it to the customer at the appropriate time.

[0411] Output: Contract concluded with customer

[0412] (Application example 3)

[0413] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0414] In conventional sales activities and customer service, it was difficult to quickly and effectively compete with new plans from other companies. There was also a lack of effective means to train new crew members and improve the efficiency of customer acquisition. Furthermore, it was difficult for staff to make optimal proposals in real time when interacting with customers.

[0415] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0416] In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and the talk of the company's excellent crew, a means for generating talk that differentiates the company from other companies when the company issues a new plan, a means for generating counter talk when other companies issue a new plan, and a means for displaying counter talk and new plan proposals in real time on smart glasses worn by the staff. This makes it possible to quickly and effectively counter new plans of other companies, and also improves the efficiency of training new staff and customer acquisition. It also makes it possible for staff to make optimal suggestions in real time when talking to customers.

[0417] "All plans and services of our company and competitors" refers to all rate plans and service contents offered by our company and competitors.

[0418] "Excellent crew talk" refers to the customer service and sales speaking style used by crew members who have particularly excellent sales records within a company.

[0419] "Artificial intelligence" refers to computer systems that mimic human intelligence using techniques such as machine learning and natural language processing.

[0420] "Differentiation talk" refers to a way of speaking that emphasizes the superiority of one's own plans or services compared to those of other companies.

[0421] "Countertalk" refers to the speech used to counter an announcement of a new plan or service by another company.

[0422] "Smart glasses" refers to a wearable device in the form of glasses that has a built-in display and sensors for displaying information.

[0423] "Real-time" refers to processing and display occurring immediately, without delay.

[0424] "New Plan Proposal" refers to a newly proposed pricing plan or service content for an existing plan or service.

[0425] The "new crew" refers to sales and customer service representatives who have recently joined the company.

[0426] "Training" refers to the education and training provided to new crew members to acquire the knowledge and skills necessary for their work.

[0427] "Talk wall" refers to a simulation in which crew members have conversations with virtual counterparts to practice their speaking style.

[0428] "Customer acquisition" refers to sales activities aimed at attracting new customers to your company's services and plans.

[0429] "Improving efficiency" refers to increasing the productivity of business or work.

[0430] The system for implementing this invention is configured as follows. First, the server trains the artificial intelligence to learn all plans and services of the company and other companies, as well as the talks of the company's top crew members. This allows the server to grasp detailed information about the plans and services of the company and other companies.

[0431] Next, the server generates differentiation talk when it launches a new plan, allowing the server to provide a way of talking to emphasize how its new plan is different from other plans.

[0432] In addition, the server generates countertalk when other companies release new plans, allowing the server to quickly provide counter talk to other companies' new plans.

[0433] The counter talk and new plan ideas generated by the server are displayed in real time on the smart glasses worn by the store clerk, allowing the clerk to instantly make optimal suggestions when interacting with the customer.

[0434] As a concrete example, consider the case where another company announces a plan that offers "unlimited calls and data for 500 yen per month." In this case, the server inputs the following prompt sentence into the generative AI model:

[0435] New plans from other companies: Unlimited calls and data for 500 yen per month

[0436] Generate counter talk to this.

[0437] Based on this prompt, the generative AI model would generate the following countertalk:

[0438] For 500 yen per month, you get unlimited calls, unlimited data, and 24 / 7 support.

[0439] This countertalk is displayed in real time on the smart glasses, allowing store associates to make quick suggestions to the customer.

[0440] The hardware used includes smart glasses (e.g., Google® Glass®), and the software uses OpenAI®'s GPT-3® API. The server links these hardware and software to generate and display countertalk in real time.

[0441] This system allows the company to quickly and effectively compete with new plans from other companies, improves the efficiency of training new staff and customer acquisition, and enables staff to make optimal suggestions in real time when interacting with customers.

[0442] The flow of the specific process in the application example 3 will be described with reference to FIG.

[0443] Step 1:

[0444] The server collects all plans and services from its own and other companies, as well as the talk of its top crew members, and trains the artificial intelligence to learn from them.

[0445] Input: Plan and service information of our company and other companies, talk data of outstanding crew members

[0446] Output: A trained artificial intelligence model

[0447] How it works: The server retrieves plan and service information and token data from the database, then inputs this into a machine learning algorithm to train the model.

[0448] Step 2:

[0449] When the server launches a new plan, it generates a differentiating feature from other companies.

[0450] Input: New plan information, trained artificial intelligence model

[0451] Output: Differentiated Talk

[0452] Specific operation: The server inputs the new plan information as a prompt into the generative AI model and generates a conversation that emphasizes the points that differentiate the company from other companies.

[0453] Step 3:

[0454] The server generates counter talk when other companies release new plans.

[0455] Input: Information on new plans from other companies, trained artificial intelligence model

[0456] Output: Countertalk

[0457] Specific operation: The server inputs information about new plans from other companies as prompts into the generative AI model, generating countertalk to counter them.

[0458] Step 4:

[0459] The server displays the generated counter talk and new plan ideas in real time on smart glasses worn by the store clerk.

[0460] Input: Countertalk, new plan

[0461] Output: Information displayed on the smart glasses

[0462] Specific operation: The server transmits the generated counter talk and new plan ideas to the smart glasses display, allowing the store clerk to view them in real time.

[0463] Step 5:

[0464] The user (store clerk) can make optimal suggestions to customers based on the counter talk and new plan ideas displayed on the smart glasses.

[0465] Input: Information displayed on smart glasses

[0466] Output: Proposal to customer

[0467] Specific operations: The user checks the information displayed on the smart glasses and makes appropriate suggestions to the customer based on that information.

[0468] Furthermore, an emotion engine that estimates the emotion of the user may be combined. That is, the identification processing unit 290 may estimate the emotion of the user using the emotion identification model 59, and perform identification processing using the emotion of the user.

[0469] "Example 1"

[0470] We will explain the form of the invention that combines an emotion engine. First, the AI ​​learns all plans and services of the company and other companies, and the talks of the company's excellent crew members. Next, the emotion engine is combined with this AI. The emotion engine has the ability to recognize the user's emotions and generate the optimal talk according to those emotions. For example, if the user is expressing joy, the emotion engine generates talk that amplifies that joy. Conversely, if the user is expressing dissatisfaction, the emotion engine generates talk that resolves that dissatisfaction. In this way, the invention that combines an emotion engine generates the optimal talk according to the user's emotions, realizing more effective sales activities.

[0471] "Example 2"

[0472] We will explain the form of the invention that combines the emotion engine with concrete examples. For example, when a company releases a new plan, the AI ​​generates a speech that differentiates it from other companies. At this time, the emotion engine recognizes the user's emotions and generates a speech that corresponds to that emotion. For example, if the user is excited, the emotion engine generates a speech that maintains that excitement. Also, if the user is feeling anxious, the emotion engine generates a speech that relieves that anxiety. In this way, the invention that combines the emotion engine generates the optimal speech according to the user's emotions, realizing more effective sales activities.

[0473] "Example 3"

[0474] The form of the invention combined with the emotion engine will be explained with concrete examples. For example, when another company releases a new plan, the AI ​​generates counter talk to that plan. At this time, the emotion engine recognizes the user's emotions and generates talk according to those emotions. For example, if the user is excited about another company's new plan, the emotion engine generates talk that calms the excitement. Also, if the user is dissatisfied with another company's new plan, the emotion engine generates talk that resolves the dissatisfaction. In this way, the invention combined with the emotion engine generates optimal talk according to the user's emotions, realizing more effective sales activities.

[0475] The process flow of each embodiment will be described below.

[0476] "Example 1"

[0477] Step 1: Have the AI ​​learn all of the plans and services of your company and other companies, as well as the talk of your company's top crew members.

[0478] Step 2: Combine the AI ​​with an emotion engine. The emotion engine has the ability to recognize the user's emotions.

[0479] Step 3: The emotion engine generates optimal speech according to the user's emotions.

[0480] "Example 2"

[0481] Step 1: When your company launches a new plan, the AI ​​generates a speech that differentiates it from other companies.

[0482] Step 2: The emotion engine recognizes the user's emotion and generates a speech according to that emotion.

[0483] Step 3: The emotion engine generates optimal conversations based on the user's emotions, enabling more effective sales activities.

[0484] "Example 3"

[0485] Step 1: When another company releases a new plan, the AI ​​creates a counter-talk to that plan.

[0486] Generate a query.

[0487] Step 2: The emotion engine recognizes the user's emotion and generates a speech according to that emotion.

[0488] Step 3: The emotion engine generates optimal conversations based on the user's emotions, enabling more effective sales activities.

[0489] Example 1

[0490] Next, a description will be given of Example 1 of the embodiment 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal".

[0491] In conventional sales activities, it was difficult to effectively utilize information on plans and services from our company and other companies to provide optimal conversations with customers. In addition, we were unable to respond appropriately to customers' emotions, making it difficult to improve customer satisfaction. Furthermore, it was difficult to train new employees and quickly respond to new plans from other companies.

[0492] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means. In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and the conversations of the company's excellent employees, a means for generating a differentiated conversation with other companies when the company issues a new plan, a means for generating a counter conversation when another company issues a new plan, a means for integrating an emotion engine, recognizing the user's emotions and generating an optimal conversation, and a means for making an artificial intelligence model learn using preprocessed data. This makes it possible to provide an optimal conversation according to the customer's emotions and improve customer satisfaction. In addition, it is also possible to realize training of new employees and a quick response to new plans of other companies.

[0493] A "plan" refers to the detailed content and terms of the products or services a company offers.

[0494] "Services" refers to various types of assistance and benefits that a company provides to its customers.

[0495] "Employees" refers to people employed by a business.

[0496] "Conversation" refers to the words and content of communication that employees have with customers.

[0497] "Artificial intelligence" refers to technology that mimics human intelligence through machine learning and data analysis.

[0498] An "emotion engine" refers to technology that recognizes a user's emotions and generates the optimal response based on those emotions.

[0499] "Preprocessing" refers to a series of operations performed to transform data into a form suitable for analysis or learning.

[0500] A "model" is a mathematical structure built on data to make predictions or classifications.

[0501] "Differentiating conversation" refers to the unique communication content that distinguishes your company's plans and services from those of other companies.

[0502] "Counter-conversations" refers to communication intended to demonstrate the superiority of one company over another company's new plans.

[0503] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0504] This invention is a system that generates optimal conversations according to customer emotions by having an artificial intelligence learn information on plans and services provided by a company and the content of conversations of outstanding employees, and integrating an emotion engine. A specific embodiment of this system will be described below.

[0505] 1. Program Generation

[0506] The server generates a program that collects information on the company's plans and services, as well as conversations between outstanding employees, and stores them in a database. This program manages the data using a database management system (e.g., MySQL (registered trademark), PostgreSQL (registered trademark)) and trains the artificial intelligence using a machine learning model (e.g., TensorFlow (registered trademark), PyTorch (registered trademark)).

[0507] 2. Program Processing

[0508] The server processes the data as follows:

[0509] 1. Data collection:

[0510] The server collects information about the company's plans and services, as well as conversations between top employees, and stores the information in a database using web scraping tools (e.g. Beautiful Soup, Scrapy) or APIs.

[0511] 2. Data preprocessing:

[0512] The server pre-processes the collected data, specifically cleaning the text data (e.g. removing unnecessary spaces and special characters), tokenizing, stemming, etc. For this purpose, it uses natural language processing libraries (e.g. NLTK, spaCy).

[0513] 3. Training the AI ​​model:

[0514] The server uses the preprocessed data to train an artificial intelligence model, using machine learning frameworks (e.g., TensorFlow (registered trademark), PyTorch (registered trademark)) and applying supervised learning and reinforcement learning techniques.

[0515] 4. Emotion engine integration:

[0516] The server integrates an emotion engine into the trained artificial intelligence model. The emotion engine recognizes the user's emotions and generates optimal conversations according to those emotions. For emotion recognition, it uses emotion analysis libraries (e.g. TextBlob, VADER).

[0517] 5. Conversation Generation:

[0518] The server generates a conversation according to the user's emotions. For example, if the user expresses a feeling of joy, the emotion engine generates a conversation that amplifies the joy. Conversely, if the user expresses a feeling of dissatisfaction, the emotion engine generates a conversation that resolves the dissatisfaction.

[0519] 3. Specific Examples

[0520] When a user inquires about a new plan, the server analyzes the user's emotions and generates an appropriate dialogue. For example, if the user is excited, the server generates a dialogue like "This plan is a great choice for you!". Conversely, if the user is dissatisfied, the server generates a dialogue like "We apologize for the inconvenience. This plan is designed to meet your needs."

[0521] Example prompt:

[0522] "Tell me about the new plan. I'm not happy with the current plan."

[0523] By inputting this prompt sentence into a generative AI model, it is possible to analyze the user's emotions and generate an appropriate conversation.

[0524] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0525] Program processing flow

[0526] Step 1: Data collection

[0527] The server collects information about the company's plans and services, as well as conversations between outstanding employees. Specifically, data is acquired using web scraping tools (e.g. Beautiful Soup, Scrapy) and APIs. The collected data is stored in a database management system (e.g. MySQL (registered trademark), PostgreSQL (registered trademark)).

[0528] Input: Data from company websites and APIs

[0529] Output: Plan and service information stored in database and conversation content

[0530] Specific behavior:

[0531] The server periodically executes a script to retrieve data from websites and APIs and store it in a database. For example, you can set it to automatically collect data every day at 2 a.m.

[0532] Step 2: Data Preprocessing

[0533] The server pre-processes the collected data, specifically cleaning the text data (e.g. removing unnecessary spaces and special characters), tokenizing, stemming, etc. For this purpose, it uses natural language processing libraries (e.g. NLTK, spaCy).

[0534] Input: Raw data stored in a database

[0535] Output: Preprocessed and clean data

[0536] Specific behavior:

[0537] The server runs a script that cleans and tokenizes the text data retrieved from the database, for example splitting the text "Hello, this is a test" into tokens such as "Hello", "This", "Is", "Test", and "Desu".

[0538] Step 3: Training the AI ​​model

[0539] The server uses the preprocessed data to train an artificial intelligence model, using machine learning frameworks (e.g., TensorFlow (registered trademark), PyTorch (registered trademark)) and applying supervised learning and reinforcement learning techniques.

[0540] Input: Preprocessed and clean data

[0541] Output: A trained artificial intelligence model

[0542] Specific behavior:

[0543] The server uses the preprocessed data to train a neural network model, for example using past conversation data as input to build a model that predicts upcoming conversations.

[0544] Step 4: Integrating the Emotion Engine

[0545] The server integrates an emotion engine into the trained artificial intelligence model. The emotion engine recognizes the user's emotions and generates optimal conversations according to those emotions. For emotion recognition, it uses emotion analysis libraries (e.g. TextBlob, VADER).

[0546] Input: Text entered by the user

[0547] Output: Sentiment score

[0548] Specific behavior:

[0549] The server passes the input text from the user to a sentiment analysis library to obtain a sentiment score. For example, it calculates the sentiment score of "joy" from the text "This plan is really great!"

[0550] Step 5: Conversation generation

[0551] The server generates a conversation according to the user's emotions. For example, if the user expresses a feeling of joy, the emotion engine generates a conversation that amplifies the joy. Conversely, if the user expresses a feeling of dissatisfaction, the emotion engine generates a conversation that resolves the dissatisfaction.

[0552] Input: Sentiment scores and a trained artificial intelligence model

[0553] Output: The optimal conversation to serve to the user

[0554] Specific behavior:

[0555] The server generates an appropriate conversation based on the emotion score obtained from the emotion engine. For example, if the user says, "This plan is really great!", the server generates a conversation such as, "Thank you! Many people are satisfied with this plan."

[0556] (Application example 1)

[0557] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0558] In conventional sales activities, it was difficult to properly recognize the customer's emotions and generate optimal conversations accordingly. In addition, there was a lack of effective means to recognize emotions and generate conversations in real time during conversations with customers. This resulted in insufficient improvement in customer satisfaction and sales efficiency. In addition, training new crew members and quickly responding to new plans from other companies were also issues.

[0559] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0560] In this invention, the server includes a means for making an AI learn all plans and services of the company and other companies, and the talk of the company's excellent crew, a means for generating talk that differentiates the company from other companies when the company issues a new plan, a means for generating counter talk when another company issues a new plan, a means for combining an emotion engine that recognizes the emotions of customers and generates optimal talk according to the emotions, and a means for analyzing the facial expressions and tone of voice of customers in real time using smart glasses and displaying the generated talk. This makes it possible to generate and display optimal talk according to the emotions of customers in real time.

[0561] "Plan / Service" refers to the content of the products and services offered by the company and their terms of use.

[0562] "Excellent crew" refers to employees within a company who have particularly outstanding performance and skills.

[0563] "AI" is an abbreviation for artificial intelligence, and refers to the technology that allows machines to imitate human intelligence and learn and reason.

[0564] An "emotion engine" refers to technology for recognizing a user's emotions and generating a response according to those emotions.

[0565] "Smart glasses" refers to a glasses-type device equipped with a display, camera, sensors, etc., and with the ability to display information and collect data.

[0566] "Facial expression analysis" refers to the technology of using devices such as cameras to recognize human facial expressions and estimate their emotions.

[0567] "Tone of voice analysis" refers to a technology that analyzes voice data to infer a speaker's emotions and intentions.

[0568] "Talk generation" refers to a technology that automatically creates appropriate responses and conversation content based on specific conditions and input data.

[0569] A system for implementing the present invention uses the following hardware and software.

[0570] Hardware

[0571] Smart Glasses

[0572] camera

[0573] server

[0574] software

[0575] OpenCV (face detection)

[0576] EmotionRecognizer

[0577] AITrainer (talk generation)

[0578] System Configuration

[0579] The server has the means to train the AI ​​to learn all plans and services of the company and other companies, as well as the talk of the company's top crew members. Furthermore, by combining it with an emotion engine, it can recognize the emotions of customers and generate the optimal talk according to those emotions.

[0580] Processing flow

[0581] 1. Face detection on camera

[0582] The camera on the smart glasses captures the customer's face in real time, and OpenCV is used to detect the face from the captured image.

[0583] 2. Emotion recognition

[0584] After face detection, EmotionRecognizer is used to recognize the emotion of the detected face, which identifies what emotion the customer is feeling.

[0585] 3. Token Creation

[0586] Based on the recognized emotions, AITrainer generates the optimal conversation, which has appropriate content according to the customer's emotions.

[0587] 4. Viewing chats

[0588] The generated conversations are displayed on the smart glasses' display, allowing store clerks to check the most appropriate conversations in real time and use them in conversations with customers.

[0589] Examples

[0590] For example, if a customer is looking for a new smartphone but is unsure which one to choose, they could input the following prompt into the generative AI model:

[0591] Examples of prompt statements

[0592] “What talk do you generate when a customer has a confused look on their face?”

[0593] Based on this prompt, AITrainer will generate the following talk:

[0594] Example of a generated token

[0595] "It seems like you're undecided. If you don't mind, I'd like to give you some recommendations."

[0596] In this way, AI combined with an emotion engine can generate and display optimal conversations in real time based on the customer's emotions, which is expected to improve customer satisfaction and sales efficiency.

[0597] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0598] Step 1:

[0599] A camera mounted on the smart glasses captures the customer's face in real time.

[0600] Input: Customer's facial video data

[0601] Output: The captured image frames

[0602] How it works: The camera continuously captures video and captures each frame as image data.

[0603] Step 2:

[0604] The server uses OpenCV to detect faces in the captured image frames.

[0605] Input: The captured image frame

[0606] Output: Face position information (coordinate data)

[0607] What it does: Apply OpenCV's face detection algorithm to locate the face within the image frame.

[0608] Step 3:

[0609] The server uses EmotionRecognizer to recognize the emotion of the detected face.

[0610] Input: Face position information (coordinate data) and face image

[0611] Output: Recognized emotion data (e.g., happy, sad, surprised, etc.)

[0612] What it does: A face image is fed into EmotionRecognizer, which applies an emotion recognition algorithm to classify the emotion.

[0613] Step 4:

[0614] The server uses AITrainer to generate optimal talk based on the recognized emotions.

[0615] Input: Recognized emotion data

[0616] Output: Generated chat (text data)

[0617] Specific operation: Emotion data is input into AITrainer, and appropriate speech is generated using a generative AI model.

[0618] Step 5:

[0619] The server displays the generated tokens on the display of the smart glasses.

[0620] Input: Generated token (text data)

[0621] Output: Talk displayed on the smart glasses display

[0622] Specific behavior: The generated token is sent to the smart glasses display and displayed for the user to view.

[0623] Step 6:

[0624] The user (store clerk) checks the chat displayed on the smart glasses and uses it in conversations with customers.

[0625] Input: Speech displayed on smart glasses

[0626] Output: Customer interaction

[0627] Specific operation: The store clerk reads the talk displayed on the smart glasses and interacts with the customer based on that.

[0628] Example 2

[0629] Next, a second embodiment of the second embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0630] In conventional sales activities, it was difficult to generate pitches to propose new plans that were differentiated from other companies' plans and services. In addition, there was a lack of ways to generate optimal pitches that matched the user's emotions, making it difficult to conduct effective sales activities. Furthermore, there was a need for efficient methods for training new employees and practicing pitches.

[0631] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and the talk of outstanding employees of the company, a means for generating talk that differentiates the company from other companies when the company issues a new plan, a means for generating counter talk when other companies issue a new plan, a means for collecting and analyzing plan information of other companies, a means for extracting differentiation points, and a means for recognizing the user's emotions and generating talk according to the emotions. This makes it possible to generate a proposal talk for a new plan that is differentiated from other companies, and to provide the optimal talk according to the user's emotions.

[0632] "Artificial intelligence" is a technology that mimics human intelligence by learning and analyzing data.

[0633] A "differentiation talk" is a proposal that emphasizes the superiority of your company's plans and services compared to those of other companies.

[0634] "Countertalk" is a proposal made to counter a new plan announced by another company.

[0635] "Plan information" is detailed information including the service content, price, conditions, etc.

[0636] "Analysis" is the process of processing collected data and understanding and classifying its contents.

[0637] A "point of differentiation" is an element that shows the characteristics or advantages of your company's plans or services that make them different from those of other companies.

[0638] "Emotion recognition" is a technology that analyzes emotions from a user's voice, facial expressions, text input, etc.

[0639] "Talk generation" is the process of automatically creating suggestions and explanations for a specific purpose.

[0640] "Employees" refers to people employed by a business.

[0641] "Training" refers to educational activities for new employees to acquire the knowledge and skills necessary for their jobs.

[0642] "Talk practice" is training for employees to effectively communicate proposals and explanations.

[0643] An "automation device" is a machine or system that performs a specific task automatically, without human assistance.

[0644] The present invention is a system for generating a proposal talk for a new plan that is differentiated from plans of other companies and providing an optimal talk according to the user's emotions. A specific embodiment of this system will be described below.

[0645] System Configuration

[0646] This system consists of three main elements: a server, a terminal, and a user. The server uses a generative AI model to collect and analyze information on plans from other companies, extracts points of differentiation, and generates a proposal for a new plan. The terminal recognizes the user's emotions and generates a proposal that corresponds to those emotions.

[0647] Hardware and Software Used

[0648] The server is a computer with high-performance computing power, and uses machine learning frameworks such as "TensorFlow (registered trademark)" and "PyTorch (registered trademark)" to execute generative AI models. It also uses web scraping technology to collect plan information from other companies. Specifically, it uses Python libraries such as "BeautifulSoup" and "Scrapy."

[0649] The terminal is a device equipped with a camera and microphone to capture the user's voice and facial expressions, and uses emotion recognition APIs such as "IBM Watson (registered trademark)" and "Microsoft (registered trademark) Azure (registered trademark) Cognitive Services."

[0650] Data processing and calculation

[0651] The server collects plan information from other companies and analyzes the content using natural language processing (NLP) technology. Based on the analysis results, the company's new plan is compared with the plans of other companies and points of differentiation are extracted. Based on the points of differentiation, the generative AI model generates a proposal speech for the new plan.

[0652] The device analyzes the user's emotions from their voice, facial expressions, text input, etc., and generates conversations according to those emotions. The emotion engine analyzes the user's emotions, and if the user is excited, it generates conversations that maintain that excitement, and if the user is anxious, it generates conversations that relieve that anxiety.

[0653] Examples

[0654] For example, when a company releases a new plan, the server collects plan information from other companies and generates differentiated talk using a generative AI model. When a user inquires about the new plan, the device analyzes the user's emotions and generates the optimal talk based on those emotions.

[0655] An example of a specific prompt is as follows:

[0656] Generate differentiated talk for your new plan based on other companies' plan information. For example, if the other company's plan is "1000 yen per month with unlimited calls," generate a differentiated talk for your new plan. Also, generate talk according to the user's emotions. Consider both scenarios when the user is excited and when he or she is anxious.

[0657] In this way, a system is realized in which the server, terminal, and user work together to generate proposal talks for new plans that are differentiated from those of other companies, and provide optimal talks that correspond to the user's emotions.

[0658] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0659] Step 1: Gather information about other companies' plans

[0660] The server collects plan information from other companies. It receives publicly available information on the Internet and database URLs as input. Specifically, the server uses web scraping technology to obtain plan information from other companies' official websites. It uses Python libraries such as "BeautifulSoup" and "Scrapy" to analyze pages containing plan information. It obtains the collected plan information as text data as output.

[0661] Step 2: Analyze other companies' plans

[0662] The server analyzes the collected plan information of other companies. As input, it receives the text data collected in step 1. Specifically, it uses natural language processing (NLP) technology to extract the plan contents and features. It analyzes the text data using "TensorFlow (registered trademark)" and "PyTorch (registered trademark)" to extract the plan price and service details. It obtains the analysis results as output.

[0663] Step 3: Extracting points of differentiation

[0664] The server compares the company's new plan with other companies' plans based on the analysis results and extracts points of differentiation. As input, it receives the analysis results obtained in step 2 and the company's plan information. Specifically, it compares the plan information of other companies with the company's plan information and lists the differences in price and service content. As output, it obtains points of differentiation.

[0665] Step 4: Generate a new plan proposal token

[0666] The server generates a proposal speech for the new plan using the generative AI model based on the differentiation points. It receives the differentiation points obtained in step 3 as input. In concrete terms, it inputs a prompt sentence into the generative AI model to generate a proposal speech for the new plan. As an example of a prompt sentence, it inputs, "If the plan of another company is 1,000 yen per month with unlimited calling charges, please generate a proposal speech for a differentiated new plan." As output, it obtains the generated proposal speech.

[0667] Step 5: Recognizing user emotions

[0668] The device recognizes the user's emotions. It receives the user's voice, facial expressions, and text input as input. Specifically, it sends data captured by the camera and microphone to an emotion recognition API. It then uses emotion recognition APIs such as "IBM Watson" and "Microsoft Azure Cognitive Services" to analyze the user's emotions. It obtains the analysis results as output.

[0669] Step 6: Generate speech based on emotions

[0670] The server generates talk according to the user's emotions based on the analysis results of the emotion engine. As input, it receives the emotion analysis results obtained in step 5 and the suggested talk generated in step 4. In concrete terms, it inputs a prompt sentence into the generative AI model and generates talk according to the emotion. As an example of a prompt sentence, it inputs "If the user is excited, please generate talk that will maintain that excitement." As output, it obtains the optimal talk according to the emotion.

[0671] (Application example 2)

[0672] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0673] In conventional sales activities, it was difficult to generate a sales pitch that differentiated the company from its competitors, and it was not possible to provide the optimal sales pitch according to the user's emotions. In particular, online shopping sites are required to provide differentiated sales pitches in real time while users are browsing products, but there was no effective way to achieve this.

[0674] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0675] In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and the talk of the company's excellent employees, a means for generating talk that differentiates the company from other companies when the company releases a new plan, a means for generating counter talk when other companies release a new plan, a means for recognizing the user's emotions and generating talk according to the emotions, a means for collecting information on other companies and learning their points of differentiation, and a means for generating optimal talk according to the user's emotions. This makes it possible to provide optimal differentiated talk according to the user's emotions in real time.

[0676] "All plans and services of our company and competitors" refers to all rate plans and service contents offered by our company and competitors.

[0677] "Top employee talk" refers to effective sales talk used by employees who are particularly successful in sales activities.

[0678] "Artificial intelligence" refers to systems that mimic human intelligence using technologies such as machine learning and natural language processing.

[0679] "Differentiation talk" refers to sales talk that emphasizes the superiority of one's company's plans or services compared to those of other companies.

[0680] "Countertalk" refers to sales pitches made to counter an announcement of a new plan or service by another company.

[0681] "Means of recognizing the user's emotions" refers to technology that uses sensors such as cameras and microphones to analyze the user's emotions from their facial expressions and tone of voice.

[0682] "Means for generating a talk according to emotions" refers to a technology for generating an optimal sales talk based on the recognized emotions of a user.

[0683] "Means of collecting information about other companies" refers to technology for collecting information about competitors' plans and services from the Internet, etc.

[0684] "Means of learning points of differentiation" refers to technology that learns in what ways a company's plans and services are superior based on collected information about other companies.

[0685] "Providing in real time" refers to generating and providing an appropriate sales pitch at the exact moment that the user is viewing a product.

[0686] As an embodiment of the present invention, the following system is constructed.

[0687] System Configuration

[0688] The system consists of a server, a user terminal (smartphone), and various software components. The server hosts an artificial intelligence (AI) model and processes data from the user terminal. The user terminal recognizes the user's emotions using a camera and microphone and transmits them to the server.

[0689] Hardware and software used

[0690] Hardware: Smartphone (camera, microphone)

[0691] Software: TensorFlow (registered trademark) (AI model), OpenCV (emotion recognition), Flask (backend), React Native (frontend)

[0692] Program Processing

[0693] 1. Data collection:

[0694] The server scrapes product information from other companies' online shopping sites and stores it in a database. This allows us to collect information about other companies' plans and services. The libraries used are BeautifulSoup and Requests.

[0695] 2. Training the AI ​​model:

[0696] The server learns the points of differentiation between the company's products and those of other companies based on the collected data. The library used is TensorFlow (registered trademark).

[0697] 3. Emotion recognition:

[0698] The user device uses a camera and microphone to analyze the user's facial expressions and tone of voice in real time. The libraries used are OpenCV and Librosa.

[0699] 4. Token creation:

[0700] The server generates differentiated speech based on the user's emotions. The library used is GPT-3 (registered trademark) (OpenAI (registered trademark)).

[0701] 5. Frontend display:

[0702] The user's device uses React Native to display the generated token to the user.

[0703] Examples

[0704] For example, if another company's plan is "unlimited calling charges for 1,000 yen per month," when the user is excited, a conversation might be generated that says, "unlimited calling charges and unlimited data charges for 500 yen per month."

[0705] Examples of prompt statements

[0706] If another company's plan is "unlimited calls for 1,000 yen per month," generate a conversation while your users are excited that says "unlimited calls and unlimited data for 500 yen per month."

[0707] In this way, it is possible to provide optimal differentiated talk in real time according to the user's emotions.

[0708] The flow of the specific process in the application example 2 will be described with reference to FIG.

[0709] Step 1:

[0710] The server scrapes product information from other companies' mail order sites and stores it in a database. Specifically, it uses BeautifulSoup and Requests to obtain information such as product names, prices, and plan details from other companies' web pages. The input is the URL of the other companies' web pages, and the output is the storage of the obtained product information in the database.

[0711] Step 2:

[0712] The server learns the points of differentiation between the company's products and those of other companies based on the collected data. Specifically, it uses TensorFlow (registered trademark) to input the collected data into an AI model and extract the points of differentiation. The input is product information stored in a database, and the output is an AI model that has learned the points of differentiation.

[0713] Step 3:

[0714] The user device uses a camera and microphone to analyze the user's facial expressions and tone of voice in real time. Specifically, it uses OpenCV and Librosa to recognize the user's emotions. The input is the video and audio data captured by the camera and microphone, and the output is the recognized user's emotion data.

[0715] Step 4:

[0716] The server generates differentiated speech based on the recognized user emotion. Specifically, it uses GPT-3 (registered trademark) (OpenAI (registered trademark)) to input the user's emotion data and differentiation points as prompts and generate optimal speech. The input is the user's emotion data and differentiation points, and the output is the generated speech.

[0717] Step 5:

[0718] The user device uses React Native to display the generated token to the user. Specifically, the generated token is displayed on the screen and provided to the user. The input is the generated token, and the output is the token displayed to the user.

[0719] Example 3

[0720] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0721] In conventional sales activities, it was difficult to generate counter talk to quickly and effectively counter new plans from other companies, or to create new plan / service proposals. In addition, when training new employees, there was often a lack of practice in effective talk. Furthermore, there were limited means to improve the efficiency of customer acquisition. To solve these problems, a system was needed that could quickly collect information on new plans from other companies, and use that information to generate effective counter talk and new plan / service proposals.

[0722] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0723] In this invention, the server includes a means for collecting new plan information from other companies, a means for creating a prompt sentence to be input to the generation AI model based on the collected information, and a means for generating counter talk using the generation AI model. This makes it possible to generate counter talk that quickly and effectively counters new plans from other companies. In addition, by including a means for creating new plan / service proposals that exploit the weaknesses of other companies' plans and services, a means for using the plan as a talking wall material during training for new employees, a means for using an emotion engine that recognizes the user's emotions and generates talk according to the emotions, and a means for sending the generated talk to a robot and for the robot to automatically conduct sales activities to customers, it becomes possible to significantly improve the efficiency and effectiveness of sales activities.

[0724] "Means of collecting information on new plans from other companies" refers to means of collecting information on new plans and services announced by other companies using web scraping tools or APIs.

[0725] The "means for creating prompt sentences to be input into the generative AI model" refers to a means for creating instruction sentences for causing the generative AI model to generate appropriate countertalk or new plan / service proposals based on collected information on new plans from other companies.

[0726] "Means for generating countertalk using a generative AI model" refers to a means for inputting a prompt sentence into a generative AI model to generate countertalk to counter a new plan of another company.

[0727] "Means for creating new plan / service proposals that exploit weaknesses in the plans and services of other companies" refers to means for generating new plan / service proposals that exploit weaknesses using a generative AI model based on detailed information about the plans and services of other companies.

[0728] "A means of using the talk generated by the generative AI model as material for bouncing talks off of during training for new employees" refers to a means of using the talk generated by the generative AI model during training for new employees to practice and simulate talks.

[0729] "Means of using an emotion engine that recognizes the user's emotions and generates talk in accordance with those emotions" refers to a means of analyzing the user's emotions with an emotion engine and, based on the results, a generative AI model generating optimal talk in accordance with the user's emotions.

[0730] "A means of sending the generated talk to a robot, which then automatically carries out sales activities toward customers" refers to a means of sending the talk generated by a generative AI model to a robot, which then automatically carries out sales activities toward customers based on that talk.

[0731] This invention is a system that generates counter talk to new plans when other companies announce new plans. The system collects information on new plans from other companies and uses a generative AI model to generate counter talk and new plan / service proposals. It can also be used to improve the efficiency of training new employees and customer acquisition.

[0732] Hardware and Software Used

[0733] Hardware

[0734] Server: Use a server with a high-performance processor and a large amount of memory. For example, use a virtual server provided by a cloud service provider (e.g., Amazon EC2, Google Cloud Compute Engine).

[0735] Device: Computers and tablets used by new employees.

[0736] Robots: Robots that automatically conduct sales activities with customers (e.g. Pepper, NAO).

[0737] software

[0738] Generative AI model: We use GPT-4 (registered trademark) from OpenAI (registered trademark) as the generative AI model.

[0739] Web scraping tools: We use BeautifulSoup and Scrapy to gather information on new plans from other companies.

[0740] Emotion Engine: Uses Affectiva and IBM Watson Tone Analyzer to recognize user emotions.

[0741] Data processing and calculation

[0742] Collecting information on new plans from other companies

[0743] The server uses web scraping tools (e.g. BeautifulSoup, Scrapy) to collect new plan information from other companies' official websites and news sites. The collected information is stored in a database.

[0744] Creating a prompt statement

[0745] The server creates a prompt sentence to be input to the generative AI model based on the collected information on new plans from other companies. As a specific example, the server generates the following prompt sentence:

[0746] "Another company has announced a plan offering unlimited calls and data for 500 yen per month. Please generate a counter-talk to this plan."

[0747] Generating counter talk

[0748] The server inputs the prompt sentence into a generative AI model (e.g., GPT-4 (registered trademark) of OpenAI (registered trademark)) to generate countertalk. The generated countertalk is stored in a database.

[0749] Creating new plans and service proposals

[0750] Based on detailed information about other companies' plans and services, the server inputs prompt sentences like the following into the generative AI model to generate new plan and service proposals.

[0751] "Generate new plan and service proposals that exploit the weaknesses of other companies' plans and services."

[0752] Training new employees

[0753] The user (new employee) uses the talk generated by the generative AI model as material for bouncing ideas off the wall. The server generates a prompt sentence like the one below and sends it to the generative AI model.

[0754] "Generate counter talks to other companies' plans for onboarding new employees."

[0755] Robot customer acquisition

[0756] The server sends the chat generated by the generative AI model to the robot. The robot automatically uses the chat to conduct sales activities with customers. As a concrete example, the following chat is generated:

[0757] "Hello, let me introduce you to our new plan. For just 500 yen per month you get unlimited calls, unlimited data, and 24 / 7 support."

[0758] Using the Emotion Engine

[0759] The server recognizes the user's emotions using an emotion engine (e.g., Affectiva, IBM Watson Tone Analyzer). Based on the analysis results, the following prompt sentences are input into the generative AI model to generate a conversation according to the user's emotions.

[0760] "If a user is unhappy with another company's new plan, generate a chat that addresses that frustration."

[0761] In this way, the system can generate counter talks and new plan / service proposals to compete with new plans from other companies, and provide optimal talks according to the user's emotions. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0762] Step 1:

[0763] The server collects new plan information from other companies. Specifically, it uses web scraping tools (e.g. BeautifulSoup, Scrapy) to obtain new plan information from other companies' official websites and news sites. As input, it uses the URLs and API endpoints of other companies' websites, and as output, it stores the new plan information in a database. The server periodically checks these sites and updates the database if there is new information.

[0764] Step 2:

[0765] The server creates a prompt sentence to be input to the generative AI model based on the collected information on new plans from other companies. It uses the new plan information stored in the database as input, and generates a prompt sentence to send to the generative AI model as output. In concrete terms, the server creates a prompt sentence such as, "Another company has announced a plan with unlimited calls and unlimited data communication for 500 yen per month. Please generate a countertalk to counter this."

[0766] Step 3:

[0767] The server inputs the prompt sentence into a generative AI model (e.g., GPT-4 (registered trademark) by OpenAI (registered trademark)) and generates countertalk. As input, it uses the prompt sentence created in step 2, and as output, it stores the generated countertalk in a database. In concrete terms, the server sends the prompt sentence to the generative AI model and receives the countertalk returned from the model.

[0768] Step 4:

[0769] The server creates a prompt sentence to be input into the generative AI model based on detailed information about other companies' plans and services, and generates new plan / service proposals. Detailed information about other companies' plans and services stored in the database is used as input, and a prompt sentence is created to generate new plan / service proposals as output. Specifically, the server generates a prompt sentence saying, "Please generate new plan / service proposals that exploit the weaknesses of other companies' plans and services."

[0770] Step 5:

[0771] The user (new employee) uses the talk generated by the generative AI model as material for bouncing ideas off the wall. The generated talk is used as input, and is provided as output as training material for the new employee. In concrete terms, the server generates a prompt message saying, "Please generate a talk that competes with other companies' plans for the training of new employees," and sends it to the generative AI model. The generated talk is provided as training material for the new employee.

[0772] Step 6:

[0773] The server sends the generated talk to the robot. The generated talk is used as input, and the robot automatically conducts sales activities to customers as output. In concrete terms, the server sends the generated talk to the robot's control system, and the robot automatically conducts sales activities to customers based on the talk.

[0774] Step 7:

[0775] The server recognizes the user's emotions using an emotion engine (e.g. Affectiva, IBM Watson Tone Analyzer). The user's voice and text data are used as input, and the user's emotional state is analyzed as output. Based on the analysis results, a prompt sentence is input into the generative AI model, and a conversation corresponding to the user's emotions is generated. In concrete terms, the server inputs the user's voice and text data into the emotion engine, and analyzes the user's emotional state. Based on the analysis results, a prompt sentence is input into the generative AI model, such as, "If the user is dissatisfied with another company's new plan, please generate a conversation that will resolve that dissatisfaction," and the generated conversation is provided to the user.

[0776] (Application example 3)

[0777] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0778] In conventional sales activities, when other companies announced new plans, it was difficult to generate a counter pitch quickly and effectively. In addition, it was not possible to provide the optimal pitch according to the customer's emotions, making it difficult to improve customer satisfaction and acquire customers efficiently. Furthermore, there was an issue of a lack of practical pitches during training for new crew members.

[0779] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0780] In this invention, the server includes a means for making the AI ​​learn all plans and services of the company and other companies, and the talk of the company's excellent crew, a means for generating talk that differentiates the company from other companies when the company issues a new plan, a means for generating counter talk when other companies issue a new plan, a means for combining an emotion engine that recognizes customer emotions and generates talk according to the emotions, and a means for displaying talk on a smart device. This allows the company to provide optimal talk according to customer emotions in real time, improving customer satisfaction and efficiently acquiring customers.

[0781] "All plans and services of our company and competitors" refers to all rate plans and service contents offered by our company and competitors.

[0782] "Excellent crew talk" refers to effective sales talk used by crew members who have particularly excellent sales results within a company.

[0783] "Means of training an AI" refers to the methods or processes for using artificial intelligence technology to train specific data or information.

[0784] "Means for generating differentiating talk" refers to a method for automatically generating talk to emphasize the superiority of a company compared to the plans or services of other companies.

[0785] "Means for generating countertalk" refers to a method for automatically generating countertalk when another company announces a new plan.

[0786] An "emotion engine" refers to technology that recognizes a user's emotions and generates an optimal response based on those emotions.

[0787] "Means for displaying talk on a smart device" refers to a method for displaying the generated talk on a device such as smart glasses or a smartphone.

[0788] As an embodiment of the present invention, a system using smart glasses in a brick-and-mortar store will be described as an example.

[0789] First, the server trains the AI ​​to learn all plans and services of the company and other companies, as well as the talk of the company's top crew members. This gives the AI ​​the ability to understand the characteristics of each plan and service and generate appropriate talk.

[0790] Next, when the server launches a new plan, it creates a differentiation pitch from other companies. For example, if another company launches a plan that offers "unlimited calls and unlimited data for 500 yen per month," the server creates a counter-pitch that says, "unlimited calls and unlimited data for 500 yen per month, plus 24 / 7 support."

[0791] In addition, the server combines an emotion engine that recognizes the customer's emotions and generates speech according to those emotions. The emotion engine analyzes the customer's facial expressions and tone of voice captured by the camera, and generates speech to calm the customer's excitement if the customer is excited, and generates speech to resolve the customer's dissatisfaction if the customer is dissatisfied.

[0792] Smart glasses are used as a means of displaying talk on smart devices. Salespeople wear the smart glasses and see counter talk generated by AI and talk according to emotions in real time while interacting with customers. This allows salespeople to instantly provide the most appropriate talk according to the customer's questions and their reactions to other companies' plans.

[0793] For example, if a customer presents a plan that offers "unlimited calls and data for 500 yen per month," the AI ​​generates a countertalk of "unlimited calls and data for 500 yen per month, plus 24 / 7 support." Furthermore, if the customer is excited about this plan, the emotion engine generates a countertalk that calms that excitement.

[0794] Examples of prompts to input to a generative AI model include:

[0795] "Another company has announced a new plan: unlimited calls and unlimited data for 500 yen per month. Please generate a counter-talk to this."

[0796] "Your customers are excited. Generate conversations that reflect this emotion: 500 yen per month for unlimited calls, unlimited data, and 24 / 7 support."

[0797] In this way, it is possible to provide the most appropriate conversation in real time based on the customer's emotions, improving customer satisfaction and efficiently acquiring new customers.

[0798] The flow of the specific process in the application example 3 will be described with reference to FIG.

[0799] Step 1:

[0800] The server trains the AI ​​to learn all plans and services from the company and other companies, as well as the conversations of the company's top crew members. Specifically, it collects plan and service information and conversation data from the database and inputs them into the AI ​​model. The AI ​​model learns this data and understands the characteristics of each plan and service. The input is plan and service information and conversation data, and the output is the trained AI model.

[0801] Step 2:

[0802] When a company issues a new plan, the server generates a differentiation token that differentiates the company from other companies. Specifically, when a user inputs new plan information, the AI ​​model compares it with other companies' plans and extracts points of differentiation. Based on this, a differentiation token is generated. The input is the new plan information, and the output is the differentiation token.

[0803] Step 3:

[0804] The server generates countertalk when another company releases a new plan. Specifically, it inputs the new plan information of the other company into the AI ​​model and generates countertalk. The input is the new plan information of the other company, and the output is the countertalk.

[0805] Step 4:

[0806] The server combines an emotion engine that recognizes the customer's emotions and generates speech corresponding to those emotions. Specifically, the customer's facial expressions and tone of voice captured by the camera are input into an emotion recognition model to analyze the customer's emotions. Based on the analysis results, the AI ​​model generates speech corresponding to the emotion. The input is the customer's facial expressions and tone of voice, and the output is speech corresponding to the emotion.

[0807] Step 5:

[0808] The terminal displays the talk on the smart device. Specifically, the terminal displays the generated talk on the display of the smart glasses. The input is the generated talk, and the output is the talk displayed on the smart glasses.

[0809] Step 6:

[0810] The user interacts with the customer based on the talk displayed through the smart glasses. Specifically, the user provides the customer with the most appropriate talk while referring to the talk displayed on the smart glasses. The input is the talk displayed on the smart glasses, and the output is the conversation with the customer.

[0811] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires a voice indicating a user input for the result of the specific processing. The control unit 46A transmits the voice data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[0812] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search<URL: https: / / OpenAI.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data according to the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0813] Another example of generative AI is Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[0814] In the above embodiment, an example was given in which the specific process was performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0815] [Second embodiment]

[0816] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0817] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0818] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a wide area network (WAN) and / or a local area network (LAN).

[0819] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0820] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs the voice according to instructions from the processor 46.

[0821] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[0822] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[0823] Fig. 4 shows an example of main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0824] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32, and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0825] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0826] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50, and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0827] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0828] "Example 1"

[0829] The present invention has a means for making AI learn all plans and services of the company and other companies, and the talk of the company's excellent crew. Specifically, information on plans and services of the company and other companies, and the contents of the talk of the company's excellent crew are stored in a database, and the AI ​​learns this information.

[0830] "Example 2"

[0831] Furthermore, when a company releases a new plan, it has the means to generate a differentiating message from other companies. Specifically, based on the information the AI ​​has learned, it generates a proposal message for a new plan that is differentiated from other companies' plans and services. For example, if another company's plan is "unlimited calling charges for 1,000 yen per month," the AI ​​generates a differentiating message that says "unlimited calling charges and unlimited data communication for 500 yen per month."

[0832] "Example 3"

[0833] In addition, it has a means to generate counter talk when other companies release new plans. Specifically, when another company releases a new plan, the AI ​​learns information about that plan and generates counter talk to counter it. For example, if another company releases a plan that offers "unlimited calls and unlimited data for 500 yen per month," the AI ​​generates the counter talk "unlimited calls and unlimited data for 500 yen per month, plus 24 / 7 support."

[0834] "Example 4"

[0835] In addition, it has the means to create new plans and service proposals that exploit the weaknesses of other companies' plans and services, and to use them as a means of bouncing ideas off new crew members during training. Specifically, based on the information that AI has learned, it creates new plans and service proposals that exploit the weaknesses of other companies' plans and services. In addition, during training of new crew members, the talks generated by AI are used as material for bouncing ideas off new crew members.

[0836] "Example 5"

[0837] Finally, in the future, we will have the means to improve the efficiency of customer acquisition by transferring the acquisition to robots. Specifically, the robot will use the chat generated by AI to improve the efficiency of customer acquisition. For example, the robot will automatically conduct sales activities to customers based on the chat generated by AI.

[0838] The process flow of each embodiment will be described below.

[0839] "Example 1"

[0840] Step 1: Accumulate a database of all plans and services from your company and other companies, as well as information on the talks of your company's top crew members.

[0841] Step 2: The AI ​​learns the information from the database.

[0842] "Example 2"

[0843] Step 1: Based on the information the AI ​​has learned, it generates a proposal speech for a new plan that is differentiated from other companies' plans and services.

[0844] Step 2: Provide the generated chat to the sales team and use it to propose new plans.

[0845] "Example 3"

[0846] Step 1: When another company releases a new plan, the AI ​​learns the information about that plan.

[0847] Step 2: Based on the information that the AI ​​has learned, it generates a counter talk to counter the new plan. Step 3: The generated counter talk is provided to the sales team and used in sales activities to counter the new plan of other companies.

[0848] "Example 4"

[0849] Step 1: Based on the information the AI ​​has learned, it creates new plan and service proposals that exploit the weaknesses of other companies' plans and services.

[0850] Step 2: When training new crew members, use the AI-generated talk as a training material.

[0851] "Example 5"

[0852] Step 1: The robot uses the AI-generated chat to improve the efficiency of customer acquisition. Step 2: Based on the AI-generated chat, the robot automatically conducts sales activities to customers.

[0853] Example 1

[0854] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[0855] With the previous system, it was difficult to effectively utilize plan and service information from the company and other companies, as well as the dialogue content of excellent employees, making it difficult to provide quick and appropriate answers to user inquiries. In addition, efficient methods were required for creating new plan and service proposals and training new employees. Furthermore, there was a lack of automated means to improve the efficiency of customer acquisition.

[0856] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0857] In this invention, the server includes a means for collecting plan and service information of the company and other companies, and the contents of dialogue of the company's excellent employees, and storing them in a database, a means for preprocessing the stored data and learning them using a generative AI model, and a means for analyzing inquiries from users and generating optimal answers using the generative AI model. This makes it possible to provide prompt and appropriate answers to inquiries from users. It also provides an efficient method for creating new plan and service proposals and training new employees. Furthermore, it is possible to realize an automation means for improving the efficiency of customer acquisition.

[0858] "Plan and service information" is detailed data on the various services and pricing plans offered by companies.

[0859] "Conversations between excellent employees" are records of actual conversations conducted by employees who were highly evaluated for their customer service.

[0860] A "database" is an information accumulation system that systematically stores collected data and enables efficient search and use.

[0861] "Preprocessing" refers to processes such as data cleaning, normalization, and tokenization to prepare collected data in a format suitable for an AI model.

[0862] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate answers to user inquiries.

[0863] A "user query" is a question or information request made by a user to the system.

[0864] "Answer generation" is the process of creating an answer based on appropriate information to a user's inquiry.

[0865] A "terminal" is a device through which a user accesses and queries the system.

[0866] "Dialogue simulation" is a training method that allows new employees to gain simulated experience interacting with real customers.

[0867] "Customer acquisition automation" is a technology that automates the process of efficiently acquiring customers.

[0868] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0869] The present invention is a system for providing prompt and appropriate responses to inquiries from users. A specific embodiment of this system will be described below.

[0870] 1. Program Generation

[0871] The server generates a program that collects the plan and service information of the company and other companies provided by the user, as well as the conversations of the company's outstanding employees, and stores them in a database. This program is developed using programming languages ​​such as Python and Java (registered trademark).

[0872] 2. Data collection and storage

[0873] The server stores the collected data in a database. A relational database such as MySQL (registered trademark) or PostgreSQL (registered trademark) is used as the database. Specifically, the following data is collected:

[0874] Detailed information about your plans and services (e.g., pricing, offerings, terms and conditions)

[0875] Detailed information about other companies' plans and services (e.g., fees, services provided, contract terms)

[0876] Conversations by your company's outstanding employees (e.g. customer service scripts, success stories)

[0877] 3. Data Preprocessing

[0878] The server preprocesses the data stored in the database, specifically cleaning the data (filling in missing values ​​and removing outliers), normalizing it, tokenizing it, etc. This prepares the data in a format suitable for the generative AI model.

[0879] 4. Training the model

[0880] The server trains a generative AI model using the preprocessed data. The AI ​​model used is a natural language processing model such as GPT-3 (registered trademark) or BERT. The learning process uses machine learning frameworks such as TensorFlow (registered trademark) or PyTorch (registered trademark).

[0881] 5. Parsing the query and generating an answer

[0882] When a user makes an inquiry through a terminal, the server analyzes the content of the inquiry. It uses natural language processing technology to understand the user's intent and extract the necessary information. The server uses a generative AI model based on the analysis results to generate a response. Specifically, it combines the optimal plan and service information and dialogue content for the user's inquiry to create a response in a natural dialogue format.

[0883] 6. Providing a response

[0884] The server sends the generated answer to the user's terminal, through which the user can check the answer and make additional inquiries if necessary.

[0885] Examples

[0886] For example, if a user asks "What is the difference between my company's plan A and another company's plan B?" the server would process it as follows:

[0887] 1. Analyze the user's query and retrieve information for Plan A and Plan B from the database.

[0888] 2. Compare the information obtained and clarify the differences.

[0889] 3. Generate answers in a natural, conversational format based on the conversations of your top employees.

[0890] Examples of prompt statements

[0891] Examples of prompts to be fed to a generative AI model include:

[0892] User: "What is the difference between my Plan A and the other company's Plan B?"

[0893] AI model: "Our company's Plan A costs 5,000 yen per month and has unlimited data. The other company's Plan B costs 4,500 yen per month and has a data limit of 10GB. In addition, Plan A includes free calls, but Plan B does not."

[0894] In this way, the server realizes a system that provides appropriate answers to user inquiries.

[0895] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0896] Step 1: Data collection

[0897] The server collects plan and service information from the company and other companies, as well as the content of conversations between the company's outstanding employees. Specifically, it uses APIs and scraping tools to obtain the necessary data from websites and internal systems. The inputs include various API endpoints and webpage URLs, and the output is the acquired plan and service information and conversation content data.

[0898] Step 2: Data accumulation

[0899] The server stores the collected data in a database. MySQL (registered trademark) or PostgreSQL (registered trademark) is used as the database. Specifically, it executes SQL queries to insert the acquired data into appropriate tables. The inputs include the collected plan and service information and conversation content data, and the output is the data stored in the database.

[0900] Step 3: Data Preprocessing

[0901] The server preprocesses the data stored in the database, specifically cleaning the data (filling in missing values ​​and removing outliers), normalizing it, tokenizing it, etc. The input is the raw data stored in the database, and the output is the preprocessed, clean data.

[0902] Step 4: Model training

[0903] The server trains a generative AI model using the preprocessed data. The AI ​​models used are GPT-3 (registered trademark) and BERT, and training is performed using TensorFlow (registered trademark) and PyTorch (registered trademark). Specifically, the data is divided into a training set and a test set, and the accuracy of the model is evaluated. The input is the preprocessed data, and the output is a trained AI model.

[0904] Step 5: Query Analysis

[0905] When a user makes a query through a terminal, the server analyzes the query. It uses natural language processing technology to understand the user's intent and extract the necessary information. The input is the query text from the user, and the output is the analyzed query.

[0906] Step 6: Answer Generation

[0907] The server generates an answer using a generative AI model based on the analysis results. Specifically, it creates an answer in a natural dialogue format by combining the optimal plan and service information and dialogue content for the user's inquiry. The input is the analyzed inquiry content and the trained AI model, and the output is the generated answer text.

[0908] Step 7: Provide your answer

[0909] The server sends the generated answer to the user's terminal, where the user can check the answer and make additional inquiries if necessary. The input is the generated answer text, and the output is the answer displayed on the user's terminal.

[0910] (Application example 1)

[0911] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[0912] With the conventional system, it was difficult to effectively utilize plan and service information from our company and other companies to make optimal proposals to users. In addition, there was a lack of a way to provide optimal advice to users through chat using AI that learned the conversation content of excellent crew members. This resulted in a lack of improvement in the user's purchasing experience, and there were concerns about a decline in competitiveness.

[0913] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0914] In this invention, the server includes a means for making an AI learn all plans and services of the company and other companies, and the talk of the company's top crew, a means for generating talk that differentiates the company from other companies when the company releases a new plan, a means for generating counter talk when other companies release a new plan, a means for proposing the best plan or service for a product that the user is considering purchasing, a means for providing the user with the best option based on the company's and other companies' plan and service information, and a means for the AI ​​that has learned the talk content of the company's top crew to provide the best advice in a chat with the user. This makes it possible to propose the best plan or service to the user and improve the purchasing experience.

[0915] "Plan / Service" refers to the content of the products and services offered by the company and their terms of use.

[0916] "Excellent crew" refers to employees within a company who have particularly outstanding performance and skills.

[0917] "Talk" refers to the way of speaking and the content of explanations used in customer interactions and sales activities.

[0918] "AI" is an abbreviation for artificial intelligence and refers to systems that mimic human intelligence using technologies such as machine learning and natural language processing.

[0919] "Differentiation talk" refers to the way of speaking and the content of explanations that emphasize the superiority of a company's plans and services compared to those of other companies.

[0920] "Countertalk" refers to the way of speaking or the explanation given when another company announces a new plan or service.

[0921] "User" refers to a customer or consumer who uses the company's plans or services.

[0922] "Chat" refers to a means of communicating in real time using text messages.

[0923] "Buying experience" refers to the overall experience and satisfaction a user feels when purchasing a product or service.

[0924] "Choices" refers to the selection of plans or services available to a User.

[0925] The system for implementing this invention has the following configuration. The server accumulates information on plans and services from the company and other companies, as well as the contents of the conversations of the company's top crew members in a database, and has the AI ​​learn this information. When a user tries to purchase a product using their smartphone, the server receives the user's input and suggests the most suitable plan and service.

[0926] Hardware and Software Used

[0927] Hardware: Smartphones, servers

[0928] Software: Python, OpenAI® API

[0929] Data processing and calculation

[0930] The server processes the data in the following steps:

[0931] 1. Building the database:

[0932] The server stores information on plans and services from the company and other companies, as well as the contents of the conversations of the company's outstanding crew members in a database, allowing the AI ​​to learn from this information.

[0933] 2. Receiving user input:

[0934] The user inputs information about the product he or she is considering purchasing using a smartphone. For example, the user might input, "I want to buy a new smartphone."

[0935] 3. Generate prompt:

[0936] The server generates prompts based on user input, including information about plans and services from the company and other companies, and the content of talks by outstanding crew members.

[0937] 4. AI-generated responses:

[0938] The server sends the generated prompts to the OpenAI® API and receives an AI-generated response, which includes recommendations for the best plan or service for the user.

[0939] 5. Display response to user:

[0940] The server then displays the response on the user's smartphone, allowing the user to select the most suitable plan or service.

[0941] Examples

[0942] If the user types "I want to buy a new smartphone", the server generates a prompt like this:

[0943] Example prompt:

[0944] What people are considering buying: A new smartphone

[0945] Please suggest the best plan and service for you.

[0946] Company Plans & Services: Plan A, Plan B, Plan C

[0947] Other companies' plans and services: Plan X, Plan Y, Plan Z

[0948] Topic of the excellent crew: This plan is very cost-effective. This service is especially recommended for beginners. This product is currently on sale and is a great deal.

[0949] Based on these prompts, the AI ​​can generate responses suggesting the best plan or service for the user, for example, "Plan A is currently on sale and is a good deal. Plan B is recommended for beginners."

[0950] This way, users can choose the best plan or service for them, improving their purchasing experience.

[0951] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0952] Step 1:

[0953] The server stores the plan and service information of the company and other companies, as well as the contents of the conversations of the company's outstanding crew members in a database. This allows the AI ​​to learn from this information. The input is the plan and service information and the contents of the conversations, and the output is a database in which this information is stored. In concrete terms, the server uses a database management system (DBMS) to organize the information and save it in an appropriate format.

[0954] Step 2:

[0955] A user uses a smartphone to input information about a product he or she is considering purchasing. For example, the input may be, "I want to buy a new smartphone." The input includes information about the product the user is considering, and the output is that information is sent to a server. Specifically, the user inputs text through a smartphone application, and the data is sent to the server.

[0956] Step 3:

[0957] The server generates a prompt based on the user's input. The prompt includes the company's and other companies' plan and service information, and the talk content of outstanding crew members. The input is the user's input information, as well as the plan and service information and talk content stored in the database, and the generated prompt is obtained as the output. In terms of specific operation, the prompt is constructed using a Python script.

[0958] Step 4:

[0959] The server sends the generated prompt to the OpenAI (registered trademark) API and obtains a response from the AI. The generated prompt is the input, and the AI's response is the output. The specific operation is to call the OpenAI (registered trademark) API, send a prompt, and receive a response.

[0960] Step 5:

[0961] The server displays the obtained response on the user's smartphone. This allows the user to select the most suitable plan and service. The input is the response from the AI, and the output is the information displayed on the user's smartphone. The specific operation is to update the UI (user interface) to display the response through the smartphone application.

[0962] In this way, users can select the most suitable plan or service, improving their purchasing experience.

[0963] Example 2

[0964] Next, a description will be given of Example 2 of the embodiment 2. In the following description, the data processing device 12 is referred to as a "server" and the smart glasses 214 are referred to as a "terminal".

[0965] In the conventional system, there was a problem that it took a lot of time and effort to generate counter talks to differentiate plans and services between one's own and those of other companies. In addition, it was difficult to quickly generate counter talks when other companies announced new plans, making it difficult to maintain competitiveness. Furthermore, there was an issue of a lack of practice in effective talks in training new employees.

[0966] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0967] In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and talk of the company's outstanding employees, a means for collecting plan and service information of other companies, a means for preprocessing the collected data, a means for inputting the preprocessed data into a generating AI model, a means for generating a differentiation talk from other companies using the generating AI model, a means for outputting the generated differentiation talk, and a means for generating a counter talk when another company issues a new plan. This makes it possible to generate differentiation talk and counter talk quickly and efficiently.

[0968] "All plans and services of our company and competitors" refers to all pricing plans and service contents offered by our company and competitors.

[0969] "Excellent employee talk" refers to the effective speaking techniques and sales pitches that a company's top employees use when dealing with customers and in sales activities.

[0970] "Artificial intelligence" refers to computer systems that mimic human intelligence using techniques such as machine learning and natural language processing.

[0971] "Means of collecting information on the plans and services of other companies" refers to methods of automatically obtaining competitors' pricing plans and service details using web scraping tools, APIs, etc.

[0972] "Means for preprocessing collected data" refers to methods for analyzing acquired data and processing it to delete unnecessary information or standardize the format.

[0973] "Means of inputting preprocessed data into a generative AI model" refers to a method of providing preprocessed data in an appropriate format to a generative AI model.

[0974] A "generative AI model" refers to an artificial intelligence model that generates new text or information based on input data.

[0975] "Means for generating differentiating talk" refers to a method of using a generative AI model to generate a speaking style or sales pitch to propose a new company's plan that is differentiated from the plans and services of other companies.

[0976] "Means for outputting the generated differentiated talk" refers to a method for providing a user with the differentiated talk created by the generative AI model.

[0977] "Means of generating countertalk" refers to the way in which you generate speech or sales pitches to counter an announcement of a new plan by another company.

[0978] The present invention is a system in which a server, a terminal, and a user work together, and is implemented as follows.

[0979] System Configuration

[0980] The server has the following main functions:

[0981] 1. Data collection function:

[0982] The server collects plan and service information from other companies. This information is obtained through web scraping tools (e.g. BeautifulSoup, Scrapy) and APIs.

[0983] 2. Data pre-processing function:

[0984] The server preprocesses the collected data, specifically cleaning the text data (removing unnecessary characters and tags) and normalizing it (unifying uppercase and lowercase letters, converting special characters).

[0985] 3. Input functions for generative AI model:

[0986] The server inputs the preprocessed data into a generative AI model (e.g., GPT-4 (registered trademark)) by providing instructions to the model using prompt sentences.

[0987] 4. Differentiated Talk Generation Function:

[0988] The generative AI model generates talk to differentiate a company's new plan based on the input information about other companies' plans.

[0989] 5. Result output function:

[0990] The server outputs the generated differentiated token and transmits it to the terminal.

[0991] Hardware and Software Used

[0992] Web scraping tools: BeautifulSoup, Scrapy

[0993] Generative AI model: GPT-4 (registered trademark)

[0994] Data preprocessing tools: Python standard library and NLP libraries (e.g. NLTK, spaCy)

[0995] Examples

[0996] Example 1:

[0997] Other companies' plan information: "Unlimited calls for 1,000 yen per month"

[0998] Generated differentiated talk: "Unlimited calls and unlimited data for just 500 yen per month."

[0999] Example prompt:

[1000] "The other company's plan is 1,000 yen per month with unlimited calls. Please generate a chat to differentiate our new plan from the others."

[1001] How it works

[1002] 1. Data collection:

[1003] The server accesses the specified URL and retrieves the HTML content. It then uses BeautifulSoup to parse the HTML and extract the necessary information (plan name, price, service details, etc.).

[1004] 2. Data preprocessing:

[1005] The server removes HTML tags from the retrieved text data, removes extra spaces and line breaks, converts all characters to lower case, and converts special characters to their standard form.

[1006] 3. Input to the generative AI model:

[1007] The server inputs the preprocessed data into a generative AI model, providing instructions to the model using prompts.

[1008] 4. Generating Differentiated Talk:

[1009] The generative AI model analyzes information about other companies' plans and generates talk to propose new, competitive plans.

[1010] 5. Result output:

[1011] The server saves the chat received from the generative AI model in text format and sends it to the user's device. The user can review the chat and use it to propose new plans for their company.

[1012] In this way, the server utilizes generative AI models to provide a system that efficiently generates differentiated talks from other companies.

[1013] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1014] Step 1:

[1015] The server collects plan and service information from other companies.

[1016] Input: Third-party website URL or API endpoint.

[1017] Specific operation: The server accesses the specified URL and retrieves the HTML content. Then, it parses the HTML using BeautifulSoup or Scrapy and extracts the necessary information (plan name, price, service content, etc.).

[1018] Output: A list of extracted plans and services from other companies.

[1019] Step 2:

[1020] The server pre-processes the collected data.

[1021] Input: Third-party plan / service information collected in step 1.

[1022] What happens: The server removes HTML tags from the retrieved text, removes extra spaces and line breaks, converts all characters to lower case, and converts special characters to their standard form.

[1023] Output: Preprocessed and clean text data.

[1024] Step 3:

[1025] The server inputs the preprocessed data into a generative AI model.

[1026] Input: The clean text data preprocessed in step 2.

[1027] Specific operation: The server inputs the preprocessed data into a generative AI model (e.g., GPT-4 (registered trademark)). At this time, it gives instructions to the model using a prompt sentence. For example, it uses a prompt sentence such as, "Competitor's plans are 1,000 yen per month with unlimited calls. In contrast, please generate a chat to differentiate our new plan."

[1028] Output: The data and prompts fed into the generative AI model.

[1029] Step 4:

[1030] The generative AI model generates talk to differentiate a company's new plan based on the input information about other companies' plans.

[1031] Input: The data and prompts fed into the generative AI model in step 3.

[1032] Specific operation: The generative AI model analyzes the plan information of other companies and generates a message to propose a new competitive plan. For example, it generates a message saying, "Unlimited calls and unlimited data for 500 yen per month."

[1033] Output: The generated differentiation talk.

[1034] Step 5:

[1035] The server outputs the generated differentiated token and transmits it to the terminal.

[1036] Input: Differentiated tokens generated in step 4.

[1037] Specific operation: The server saves the chat received from the generative AI model in text format and sends it to the user's device. The user can review the chat and use it to propose new plans for their company.

[1038] Output: Differentiated token sent to user's device.

[1039] (Application example 2)

[1040] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server", and the smart glasses 214 will be referred to as a "terminal".

[1041] In conventional systems, it is difficult to generate proposal talks for new plans to differentiate from competitors, and there is a lack of means for operators of online shopping sites to effectively differentiate from competitors. In addition, there is a lack of a way to easily share the generated proposal talks, which makes it difficult to respond quickly.

[1042] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1043] In this invention, the server includes a means for making an AI learn all plans and services of the company and other companies, and talks of the company's excellent crew, a means for generating talks that differentiate the company from other companies when the company issues a new plan, a means for generating counter talk when another company issues a new plan, a means for receiving a competitor's plan as an input and generating a proposal talk for the differentiated new plan, and a means for displaying the generated proposal talk to users and making it easy to share. This makes it possible to quickly and effectively generate and share proposal talks for new plans to differentiate the company from competitors.

[1044] "Plan" refers to the content and pricing of services or products offered by a company.

[1045] "Service" refers to any intangible value or benefit provided to a customer.

[1046] "Talk" refers to explanations or proposals given to customers in sales and marketing.

[1047] "AI" is an abbreviation for artificial intelligence, and refers to the technology that allows machines to imitate human intelligence to learn and make inferences.

[1048] "Differentiation talk" refers to explanations and proposals that emphasize the advantages of your company compared to your competitors' plans and services.

[1049] "Countertalk" refers to explanations or proposals made to counter a new plan announced by a competitor.

[1050] "Competitors" refers to other companies in the same market that offer similar plans or services.

[1051] A "proposal talk" refers to an explanation or suggestion given when proposing a new plan or service to a customer.

[1052] "User" refers to individuals or companies that use this system.

[1053] "Sharing" refers to sharing generated information or data with other users.

[1054] A system for implementing the present invention includes a server, a user terminal, and a generative AI model. The server receives input from the user terminal and generates differentiated talk and counter talk using the generative AI model. Specific embodiments are described below.

[1055] System Configuration

[1056] 1. Server

[1057] The server includes a database, an AI model, and an API. The database stores the plans and services of the company and other companies, and the talks of outstanding crew members. The AI ​​model is used to learn from these data and generate differentiated talks and counter talks. The API provides an interface for communication with user terminals.

[1058] 2. User terminal

[1059] The user terminal is a device such as a smartphone or a PC that is used by the user to input competitors' plans and check the generated tokens. The user terminal communicates with the server to send and receive the necessary data.

[1060] 3. Generative AI Models

[1061] Generative AI models use advanced natural language processing models such as OpenAI's GPT-3, which generate differentiation and countertalk based on a prompt.

[1062] Processing flow

[1063] 1. Data Entry

[1064] The user inputs a competitor's plan using the user terminal. For example, the user inputs a plan that offers "unlimited calls for 1,000 yen per month."

[1065] 2. Prompt generation

[1066] The server generates a prompt based on the input plan. An example of a prompt is "If the plan of another company is '1000 yen per month with unlimited calls,' please generate a proposal for a new differentiated plan."

[1067] 3. Creating a token

[1068] The server sends a prompt to the generative AI model to generate differentiated conversations. For example, the conversations generated might be something like, "Unlimited calls and unlimited data for 500 yen per month."

[1069] 4. View and share your results

[1070] The generated token is sent to the user's terminal and displayed to the user, who can then easily share the token.

[1071] Hardware and Software Used

[1072] Hardware: Servers and user devices (smartphones, PCs, etc.) that can connect to the Internet

[1073] Software: Python, OpenAI® API, database management system (e.g. MySQL®)

[1074] Examples

[1075] If the user enters a competitor's plan, "Unlimited calls for 1000 yen per month," the server generates the following prompt:

[1076] If another company's plan offers unlimited calls for 1,000 yen per month, create a pitch for a new, differentiated plan.

[1077] Based on this prompt, the generative AI model generates a differentiated conversation that says, "Unlimited calls and unlimited data for 500 yen per month," and displays it on the user's device.

[1078] In this way, users can quickly generate and share effective pitch talks that differentiate them from their competitors.

[1079] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1080] Step 1:

[1081] User enters competitor's plan.

[1082] Input: Competitor's plan (e.g. "Unlimited calls for 1000 yen per month")

[1083] Output: The input plan data

[1084] Specific operation: The user uses a user device such as a smartphone or a PC to input the competitor's plan into an input form. The input data is then sent to the server.

[1085] Step 2:

[1086] The server receives the entered plan data and generates a prompt sentence.

[1087] Input: Plan data entered by the user

[1088] Output: Generated prompt text (e.g. "If the other company's plan is '1000 yen per month with unlimited calls,' please generate a proposal for a new differentiated plan.")

[1089] Specific operation: The server generates a prompt sentence to be sent to the generative AI model based on the received plan data. The prompt sentence includes instructions for generating a differentiation speech based on the competitor's plan.

[1090] Step 3:

[1091] The server sends prompts to the generative AI model to generate differentiated speech.

[1092] Input: Generated prompt text

[1093] Output: Generated differentiated talk (e.g. "Unlimited calls and unlimited data for 500 yen per month")

[1094] Specific operation: The server sends the generated prompt sentence to the generative AI model through the OpenAI (registered trademark) API. The generative AI model generates differentiated speech based on the prompt sentence and returns the result to the server.

[1095] Step 4:

[1096] The server transmits the generated differentiated token to the user terminal.

[1097] Input: Generated differentiation token

[1098] Output: Differentiated tokens displayed on the user's device

[1099] Specific operation: The server sends the differentiated token received from the generative AI model to the user terminal, which displays the token to the user.

[1100] Step 5:

[1101] Users can review the generated differentiated talks and share them if desired.

[1102] Input: Differentiated tokens displayed on the user's device

[1103] Output: Shared differentiation talk (e.g. share via email or social media)

[1104] Specific operation: The user checks the differentiated talk displayed on the user terminal and shares it with other users or related parties as necessary. Sharing is done through means such as email or SNS.

[1105] In this way, users can quickly generate and share effective pitch talks that differentiate them from their competitors.

[1106] Example 3

[1107] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[1108] In the conventional system, it was difficult to generate counter talk quickly and effectively when other companies announced new plans. In addition, there was a lack of means to generate and utilize effective talks when training new crew members and improving the efficiency of customer acquisition. This led to concerns about a decline in competitiveness and a decline in customer satisfaction.

[1109] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1110] In this invention, the server includes a means for collecting information on new plans of other companies, a means for making an artificial intelligence learn the collected information, and a means for storing the generated counter talk in a database. This makes it possible to generate counter talk that quickly and effectively counters new plans of other companies. In addition, by including a means for using the talk generated during training of new crew members as material for bouncing ideas off each other, and a means for a robot to use the talk generated by the artificial intelligence to improve the efficiency of customer acquisition, it is possible to improve the training effect of new crew members and to improve the efficiency of customer acquisition.

[1111] "Artificial intelligence" is a computer system that has the ability to learn from data, recognize patterns, and generate new information.

[1112] "Countertalk" is persuasive dialogue generated to counter other companies' plans or services.

[1113] A "database" is a system for efficiently storing, managing, and retrieving information.

[1114] "New crew" refers to newly hired employees or staff.

[1115] "Beat the Wall" is a simulation technique used as an interlocutor during dialogue or training.

[1116] A "robot" is a mechanical device that acts automatically according to programmed instructions to perform specific tasks.

[1117] "Customer acquisition" is the activity of attracting new customers and encouraging them to sign contracts or make purchases.

[1118] A "plan" is a contract that specifies the terms and conditions for providing a specific service or product.

[1119] "Service Proposal" means a proposal outlining the specific content and conditions of the services to be provided.

[1120] "Collection" is the act of gathering necessary information or data.

[1121] "Learning" is the process of analyzing data and acquiring patterns and knowledge.

[1122] "Generation" is the act of producing new information or data.

[1123] This invention is a system that generates counter talks quickly and effectively when other companies announce new plans. It also contributes to training new crew members and improving the efficiency of customer acquisition.

[1124] Hardware and Software Used

[1125] Hardware: Servers, terminals, robots

[1126] Software: Generative AI models (e.g., GPT-4®), database management systems (e.g., MySQL®), natural language processing tools (e.g., spaCy)

[1127] Specific implementation of the system

[1128] 1. Collect information on new plans from other companies

[1129] The server uses a web scraping tool (e.g. BeautifulSoup) to collect information on new plans from other companies' websites. Specifically, the server periodically checks other companies' websites, and if a new plan is announced, the information is stored in the database.

[1130] Specific behavior: The server scrapes the page "https: / / example.com / new-plans", extracts the new plan information, and saves it to the database.

[1131] 2. Learning Information

[1132] The server inputs the new plan information collected by the server into a generative AI model (e.g., GPT-4 (registered trademark)) and trains it. The server obtains new plan information from the database and feeds it to the generative AI model.

[1133] Specific operation: The server retrieves plan information for "unlimited calls and data for 500 yen per month" from the database and inputs it into the generative AI model.

[1134] 3. Generating countertalk

[1135] The server uses generative AI models to generate counter talk to counter other companies' plans, which are then stored in a database for future use.

[1136] Specific operation: The server inputs a prompt to the generative AI model: "Another company has announced a plan that offers unlimited calls and unlimited data for 500 yen per month. Please generate countertalk to this plan." and stores the generated countertalk in a database.

[1137] 4. Creating new plans and service proposals

[1138] The server uses the generative AI model to create new plans and service proposals that exploit the weaknesses of other companies' plans. The generated new plans and service proposals are also stored in the database.

[1139] Specific operation: The server inputs a prompt to the generative AI model saying, "Other companies' plans have a weakness in that their support systems are insufficient. Please propose a new plan that addresses this weakness." and saves the new plan generated in the database.

[1140] 5. Support for training new crew members

[1141] When training new crew members, the terminals will use the AI-generated chats as material for them to bounce ideas off of. New crew members will communicate with the AI ​​through the terminals and receive practical training.

[1142] What it does: A new crew member uses a terminal to type the prompt, "Please generate a talk that explains the benefits of our new plan," and then practices the generated talk.

[1143] 6. Improved customer acquisition efficiency

[1144] The robot uses the AI-generated talk to automatically conduct sales activities with customers. The robot makes a presentation to the customer based on the generated talk, promoting contracts.

[1145] Specific operation: The robot makes sales calls to customers using the phrase, "For 500 yen per month, you get unlimited calls, unlimited data, and 24 / 7 support."

[1146] Examples and prompts

[1147] Example 1: Generating countertalk

[1148] Other companies' plans: "Unlimited calls and data for 500 yen per month"

[1149] Generated countertalk: "For just ¥500 a month, you get unlimited calls, unlimited data, and 24 / 7 support."

[1150] Examples of prompt statements

[1151] Another company has announced a plan that offers unlimited calls and data for 500 yen per month. Please generate a counter-talk to this plan.

[1152] Example 2: Creating a new plan / service proposal

[1153] Weakness of other companies' plans: Insufficient support system

[1154] The new plan created: "Unlimited calls and data for 600 yen per month, plus 24 / 7 support and chat support via a dedicated app."

[1155] Examples of prompt statements

[1156] The weak point of other companies' plans is the insufficient support system. Please propose a new plan that takes advantage of this weak point.

[1157] Example 3: Support for new crew training

[1158] Generated chat: "Dear customer, our new plan includes 24 / 7 support, so you can use our services with peace of mind."

[1159] Examples of prompt statements

[1160] Generate onboarding talks for new crew members explaining the benefits of our new plan.

[1161] In this way, the system performs specific processing at each step, generates counter talks and new plan / service proposals to compete with new plans from other companies, and helps improve the efficiency of training new crew members and customer acquisition. The flow of the specific processing in the third embodiment will be described with reference to FIG.

[1162] Step 1:

[1163] Collecting information on new plans from other companies

[1164] The server uses a web scraping tool (e.g. BeautifulSoup) to collect information about new plans from other companies' websites.

[1165] Input: Third-party website URL

[1166] Specific action: The server scrapes the page "https: / / example.com / new-plans" and extracts information about new plans.

[1167] Data processing: Convert the extracted information into structured data (e.g., JSON format).

[1168] Output: Structured new plan information

[1169] Step 2:

[1170] Learning Information

[1171] The information about new plans collected by the server is input into a generative AI model (e.g., GPT-4 (registered trademark)) and trained.

[1172] Input: Structured new plan information

[1173] Specific operation: The server retrieves plan information for "unlimited calls and data for 500 yen per month" from the database and inputs it into the generative AI model.

[1174] Data computation: Generative AI models analyze new plan information and learn features and benefits.

[1175] Output: A trained generative AI model

[1176] Step 3:

[1177] Generating counter talk

[1178] The server uses generative AI models to generate counter talk to rival plans.

[1179] Input: trained generative AI model, prompt

[1180] Specific operation: The server inputs a prompt to the generated AI model: "Another company has announced a plan that offers unlimited calls and unlimited data for 500 yen per month. Please generate a countertalk to this plan."

[1181] Data computation: A generative AI model analyzes prompts and generates countertalk.

[1182] Output: The generated countertalk

[1183] Step 4:

[1184] Saving Countertalk

[1185] The server stores the generated countertalk in a database.

[1186] Input: Generated countertalk

[1187] Specific behavior: The server stores the generated countertalk in a database.

[1188] Data processing: Converting countertalk into an appropriate format (e.g., text format).

[1189] Output: Countertalk stored in the database

[1190] Step 5:

[1191] Creating new plans and service proposals

[1192] The server uses the generative AI model to create new plans and service proposals that exploit the weaknesses of other companies' plans.

[1193] Input: trained generative AI model, prompt

[1194] Specific operation: The server inputs a prompt to the generative AI model: "Other companies' plans have a weakness in that their support systems are insufficient. Please propose a new plan that addresses this weakness."

[1195] Data calculation: A generative AI model analyzes the prompt text and generates new plans and service proposals.

[1196] Output: Generated new plans and service proposals

[1197] Step 6:

[1198] Support for training new crew members

[1199] When training new crew members, the device will use the AI-generated talk as material for training.

[1200] Input: Generated chat, prompt

[1201] Specific Action: A new crew member uses a terminal to type the prompt, "Please generate a talk that explains the benefits of our new plan."

[1202] Data computation: The generative AI model analyzes the prompt and generates the talk.

[1203] Output: The generated token

[1204] Step 7:

[1205] Improved customer acquisition efficiency

[1206] The robot will use AI-generated chatter to automatically conduct sales activities with customers.

[1207] Input: The generated token

[1208] Specific operation: The robot makes sales calls to customers using the phrase, "For 500 yen per month, you get unlimited calls, unlimited data, and 24 / 7 support."

[1209] Data calculation: The robot analyzes the talk and presents it to the customer at the appropriate time.

[1210] Output: Contract concluded with customer

[1211] (Application example 3)

[1212] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server", and the smart glasses 214 will be referred to as a "terminal".

[1213] In conventional sales activities and customer service, it was difficult to quickly and effectively compete with new plans from other companies. There was also a lack of effective means to train new crew members and improve the efficiency of customer acquisition. Furthermore, it was difficult for staff to make optimal proposals in real time when interacting with customers.

[1214] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1215] In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and the talk of the company's excellent crew, a means for generating talk that differentiates the company from other companies when the company issues a new plan, a means for generating counter talk when other companies issue a new plan, and a means for displaying counter talk and new plan proposals in real time on smart glasses worn by the staff. This makes it possible to quickly and effectively counter new plans of other companies, and also improves the efficiency of training new staff and customer acquisition. It also makes it possible for staff to make optimal suggestions in real time when talking to customers.

[1216] "All plans and services of our company and competitors" refers to all rate plans and service contents offered by our company and competitors.

[1217] "Excellent crew talk" refers to the customer service and sales speaking style used by crew members who have particularly excellent sales records within a company.

[1218] "Artificial intelligence" refers to computer systems that mimic human intelligence using techniques such as machine learning and natural language processing.

[1219] "Differentiation talk" refers to a way of speaking that emphasizes the superiority of one's own plans or services compared to those of other companies.

[1220] "Countertalk" refers to the speech used to counter an announcement of a new plan or service by another company.

[1221] "Smart glasses" refers to a wearable device in the form of glasses that has a built-in display and sensors for displaying information.

[1222] "Real-time" refers to processing and display occurring immediately, without delay.

[1223] "New Plan Proposal" refers to a newly proposed pricing plan or service content for an existing plan or service.

[1224] The "new crew" refers to sales and customer service representatives who have recently joined the company.

[1225] "Training" refers to the education and training provided to new crew members to acquire the knowledge and skills necessary for their work.

[1226] "Talk wall" refers to a simulation in which crew members have conversations with virtual counterparts to practice their speaking style.

[1227] "Customer acquisition" refers to sales activities aimed at attracting new customers to your company's services and plans.

[1228] "Improving efficiency" refers to increasing the productivity of business or work.

[1229] The system for implementing this invention is configured as follows. First, the server trains the artificial intelligence to learn all plans and services of the company and other companies, as well as the talks of the company's top crew members. This allows the server to grasp detailed information about the plans and services of the company and other companies.

[1230] Next, the server generates differentiation talk when it launches a new plan, allowing the server to provide a way of talking to emphasize how its new plan is different from other plans.

[1231] In addition, the server generates countertalk when other companies release new plans, allowing the server to quickly provide counter talk to other companies' new plans.

[1232] The counter talk and new plan ideas generated by the server are displayed in real time on the smart glasses worn by the store clerk, allowing the clerk to instantly make optimal suggestions when interacting with the customer.

[1233] As a concrete example, consider the case where another company announces a plan that offers "unlimited calls and data for 500 yen per month." In this case, the server inputs the following prompt sentence into the generative AI model:

[1234] New plans from other companies: Unlimited calls and data for 500 yen per month

[1235] Generate counter talk to this.

[1236] Based on this prompt, the generative AI model would generate the following countertalk:

[1237] For 500 yen per month, you get unlimited calls, unlimited data, and 24 / 7 support.

[1238] This countertalk is displayed in real time on the smart glasses, allowing store associates to make quick suggestions to the customer.

[1239] The hardware used includes smart glasses (e.g., Google Glass), and the software uses OpenAI's GPT-3 API. The server links these hardware and software together to generate and display countertalk in real time.

[1240] This system allows the company to quickly and effectively compete with new plans from other companies, improves the efficiency of training new staff and customer acquisition, and enables staff to make optimal suggestions in real time when interacting with customers.

[1241] The flow of the specific process in the application example 3 will be described with reference to FIG.

[1242] Step 1:

[1243] The server collects all plans and services from its own and other companies, as well as the talk of its top crew members, and trains the artificial intelligence to learn from them.

[1244] Input: Plan and service information of our company and other companies, talk data of outstanding crew members

[1245] Output: A trained artificial intelligence model

[1246] How it works: The server retrieves plan and service information and token data from the database, then inputs this into a machine learning algorithm to train the model.

[1247] Step 2:

[1248] When the server launches a new plan, it generates a differentiating feature from other companies.

[1249] Input: New plan information, trained artificial intelligence model

[1250] Output: Differentiated Talk

[1251] Specific operation: The server inputs the new plan information as a prompt into the generative AI model and generates a conversation that emphasizes the points that differentiate the company from other companies.

[1252] Step 3:

[1253] The server generates counter talk when other companies release new plans.

[1254] Input: Information on new plans from other companies, trained artificial intelligence model

[1255] Output: Countertalk

[1256] Specific operation: The server inputs information about new plans from other companies as prompts into the generative AI model, generating countertalk to counter them.

[1257] Step 4:

[1258] The server displays the generated counter talk and new plan ideas in real time on smart glasses worn by the store clerk.

[1259] Input: Countertalk, new plan

[1260] Output: Information displayed on the smart glasses

[1261] Specific operation: The server transmits the generated counter talk and new plan ideas to the smart glasses display, allowing the store clerk to view them in real time.

[1262] Step 5:

[1263] The user (store clerk) can make optimal suggestions to customers based on the counter talk and new plan ideas displayed on the smart glasses.

[1264] Input: Information displayed on smart glasses

[1265] Output: Proposal to customer

[1266] Specific operations: The user checks the information displayed on the smart glasses and makes appropriate suggestions to the customer based on that information.

[1267] In addition, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1268] "Example 1"

[1269] We will explain the form of the invention that combines an emotion engine. First, the AI ​​learns all plans and services of the company and other companies, and the talks of the company's excellent crew members. Next, the emotion engine is combined with this AI. The emotion engine has the ability to recognize the user's emotions and generate the optimal talk according to those emotions. For example, if the user is expressing joy, the emotion engine generates talk that amplifies that joy. Conversely, if the user is expressing dissatisfaction, the emotion engine generates talk that resolves that dissatisfaction. In this way, the invention that combines an emotion engine generates the optimal talk according to the user's emotions, realizing more effective sales activities.

[1270] "Example 2"

[1271] We will explain the form of the invention that combines the emotion engine with concrete examples. For example, when a company releases a new plan, the AI ​​generates a speech that differentiates it from other companies. At this time, the emotion engine recognizes the user's emotions and generates a speech that corresponds to that emotion. For example, if the user is excited, the emotion engine generates a speech that maintains that excitement. Also, if the user is feeling anxious, the emotion engine generates a speech that relieves that anxiety. In this way, the invention that combines the emotion engine generates the optimal speech according to the user's emotions, realizing more effective sales activities.

[1272] "Example 3"

[1273] The form of the invention combined with the emotion engine will be explained with concrete examples. For example, when another company releases a new plan, the AI ​​generates counter talk to that plan. At this time, the emotion engine recognizes the user's emotions and generates talk according to those emotions. For example, if the user is excited about another company's new plan, the emotion engine generates talk that calms the excitement. Also, if the user is dissatisfied with another company's new plan, the emotion engine generates talk that resolves the dissatisfaction. In this way, the invention combined with the emotion engine generates optimal talk according to the user's emotions, realizing more effective sales activities.

[1274] The process flow of each embodiment will be described below.

[1275] "Example 1"

[1276] Step 1: Have the AI ​​learn all of the plans and services of your company and other companies, as well as the talk of your company's top crew members.

[1277] Step 2: Combine the AI ​​with an emotion engine. The emotion engine has the ability to recognize the user's emotions.

[1278] Step 3: The emotion engine generates optimal speech according to the user's emotions.

[1279] "Example 2"

[1280] Step 1: When your company launches a new plan, the AI ​​generates a speech that differentiates it from other companies.

[1281] Step 2: The emotion engine recognizes the user's emotion and generates a speech according to that emotion.

[1282] Step 3: The emotion engine generates optimal conversations based on the user's emotions, enabling more effective sales activities.

[1283] "Example 3"

[1284] Step 1: When another company puts out a new plan, the AI ​​generates a counter-talk to that plan.

[1285] Step 2: The emotion engine recognizes the user's emotion and generates a speech according to that emotion.

[1286] Step 3: The emotion engine generates optimal conversations based on the user's emotions, enabling more effective sales activities.

[1287] Example 1

[1288] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[1289] In conventional sales activities, it was difficult to effectively utilize information on plans and services from our company and other companies to provide optimal conversations with customers. In addition, we were unable to respond appropriately to customers' emotions, making it difficult to improve customer satisfaction. Furthermore, it was difficult to train new employees and quickly respond to new plans from other companies.

[1290] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means. In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and the conversations of the company's excellent employees, a means for generating a differentiated conversation with other companies when the company issues a new plan, a means for generating a counter conversation when another company issues a new plan, a means for integrating an emotion engine, recognizing the user's emotions and generating an optimal conversation, and a means for making an artificial intelligence model learn using preprocessed data. This makes it possible to provide an optimal conversation according to the customer's emotions and improve customer satisfaction. In addition, it is also possible to realize training of new employees and a quick response to new plans of other companies.

[1291] A "plan" refers to the detailed content and terms of the products or services a company offers.

[1292] "Services" refers to various types of assistance and benefits that a company provides to its customers.

[1293] "Employees" refers to people employed by a business.

[1294] "Conversation" refers to the words and content of communication that employees have with customers.

[1295] "Artificial intelligence" refers to technology that mimics human intelligence through machine learning and data analysis.

[1296] An "emotion engine" refers to technology that recognizes a user's emotions and generates the optimal response based on those emotions.

[1297] "Preprocessing" refers to a series of operations performed to transform data into a form suitable for analysis or learning.

[1298] A "model" is a mathematical structure built on data to make predictions or classifications.

[1299] "Differentiating conversation" refers to the unique communication content that distinguishes your company's plans and services from those of other companies.

[1300] "Counter-conversations" refers to communication intended to demonstrate the superiority of one company over another company's new plans.

[1301] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[1302] This invention is a system that generates optimal conversations according to customer emotions by having an artificial intelligence learn information on plans and services provided by a company and the content of conversations of outstanding employees, and integrating an emotion engine. A specific embodiment of this system will be described below.

[1303] 1. Program Generation

[1304] The server generates a program that collects information on the company's plans and services, as well as conversations between outstanding employees, and stores them in a database. This program manages the data using a database management system (e.g., MySQL (registered trademark), PostgreSQL (registered trademark)) and trains the artificial intelligence using a machine learning model (e.g., TensorFlow (registered trademark), PyTorch (registered trademark)).

[1305] 2. Program Processing

[1306] The server processes the data as follows:

[1307] 1. Data collection:

[1308] The server collects information about the company's plans and services, as well as conversations between top employees, and stores the information in a database using web scraping tools (e.g. Beautiful Soup, Scrapy) or APIs.

[1309] 2. Data preprocessing:

[1310] The server pre-processes the collected data, specifically cleaning the text data (e.g. removing unnecessary spaces and special characters), tokenizing, stemming, etc. For this purpose, it uses natural language processing libraries (e.g. NLTK, spaCy).

[1311] 3. Training the AI ​​model:

[1312] The server uses the preprocessed data to train an artificial intelligence model, using machine learning frameworks (e.g., TensorFlow (registered trademark), PyTorch (registered trademark)) and applying supervised learning and reinforcement learning techniques.

[1313] 4. Emotion engine integration:

[1314] The server integrates an emotion engine into the trained artificial intelligence model. The emotion engine recognizes the user's emotions and generates optimal conversations according to those emotions. For emotion recognition, it uses emotion analysis libraries (e.g. TextBlob, VADER).

[1315] 5. Conversation Generation:

[1316] The server generates a conversation according to the user's emotions. For example, if the user expresses a feeling of joy, the emotion engine generates a conversation that amplifies the joy. Conversely, if the user expresses a feeling of dissatisfaction, the emotion engine generates a conversation that resolves the dissatisfaction.

[1317] 3. Specific Examples

[1318] When a user inquires about a new plan, the server analyzes the user's emotions and generates an appropriate dialogue. For example, if the user is excited, the server generates a dialogue like "This plan is a great choice for you!". Conversely, if the user is dissatisfied, the server generates a dialogue like "We apologize for the inconvenience. This plan is designed to meet your needs."

[1319] Example prompt:

[1320] "Tell me about the new plan. I'm not happy with the current plan."

[1321] By inputting this prompt sentence into a generative AI model, it is possible to analyze the user's emotions and generate an appropriate conversation.

[1322] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1323] Program processing flow

[1324] Step 1: Data collection

[1325] The server collects information about the company's plans and services, as well as conversations between outstanding employees. Specifically, data is acquired using web scraping tools (e.g. Beautiful Soup, Scrapy) and APIs. The collected data is stored in a database management system (e.g. MySQL (registered trademark), PostgreSQL (registered trademark)).

[1326] Input: Data from company websites and APIs

[1327] Output: Plan and service information stored in database and conversation content

[1328] Specific behavior:

[1329] The server periodically executes a script to retrieve data from websites and APIs and store it in a database. For example, you can set it to automatically collect data every day at 2 a.m.

[1330] Step 2: Data Preprocessing

[1331] The server pre-processes the collected data, specifically cleaning the text data (e.g. removing unnecessary spaces and special characters), tokenizing, stemming, etc. For this purpose, it uses natural language processing libraries (e.g. NLTK, spaCy).

[1332] Input: Raw data stored in a database

[1333] Output: Preprocessed and clean data

[1334] Specific behavior:

[1335] The server runs a script that cleans and tokenizes the text data retrieved from the database, for example splitting the text "Hello, this is a test" into tokens such as "Hello", "This", "Is", "Test", and "Desu".

[1336] Step 3: Training the AI ​​model

[1337] The server uses the preprocessed data to train an artificial intelligence model, using machine learning frameworks (e.g., TensorFlow (registered trademark), PyTorch (registered trademark)) and applying supervised learning and reinforcement learning techniques.

[1338] Input: Preprocessed and clean data

[1339] Output: A trained artificial intelligence model

[1340] Specific behavior:

[1341] The server uses the preprocessed data to train a neural network model, for example using past conversation data as input to build a model that predicts upcoming conversations.

[1342] Step 4: Integrating the Emotion Engine

[1343] The server integrates an emotion engine into the trained artificial intelligence model. The emotion engine recognizes the user's emotions and generates optimal conversations according to those emotions. For emotion recognition, it uses emotion analysis libraries (e.g. TextBlob, VADER).

[1344] Input: Text entered by the user

[1345] Output: Sentiment score

[1346] Specific behavior:

[1347] The server passes the input text from the user to a sentiment analysis library to obtain a sentiment score. For example, it calculates the sentiment score of "joy" from the text "This plan is really great!"

[1348] Step 5: Conversation generation

[1349] The server generates a conversation according to the user's emotions. For example, if the user expresses a feeling of joy, the emotion engine generates a conversation that amplifies the joy. Conversely, if the user expresses a feeling of dissatisfaction, the emotion engine generates a conversation that resolves the dissatisfaction.

[1350] Input: Sentiment scores and a trained artificial intelligence model

[1351] Output: The optimal conversation to serve to the user

[1352] Specific behavior:

[1353] The server generates an appropriate conversation based on the emotion score obtained from the emotion engine. For example, if the user says, "This plan is really great!", the server generates a conversation such as, "Thank you! Many people are satisfied with this plan."

[1354] (Application example 1)

[1355] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[1356] In conventional sales activities, it was difficult to properly recognize the customer's emotions and generate optimal conversations accordingly. In addition, there was a lack of effective means to recognize emotions and generate conversations in real time during conversations with customers. This resulted in insufficient improvement in customer satisfaction and sales efficiency. In addition, training new crew members and quickly responding to new plans from other companies were also issues.

[1357] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1358] In this invention, the server includes a means for making an AI learn all plans and services of the company and other companies, and the talk of the company's excellent crew, a means for generating talk that differentiates the company from other companies when the company issues a new plan, a means for generating counter talk when another company issues a new plan, a means for combining an emotion engine that recognizes the emotions of customers and generates optimal talk according to the emotions, and a means for analyzing the facial expressions and tone of voice of customers in real time using smart glasses and displaying the generated talk. This makes it possible to generate and display optimal talk according to the emotions of customers in real time.

[1359] "Plan / Service" refers to the content of the products and services offered by the company and their terms of use.

[1360] "Excellent crew" refers to employees within a company who have particularly outstanding performance and skills.

[1361] "AI" is an abbreviation for artificial intelligence, and refers to the technology that allows machines to imitate human intelligence and learn and reason.

[1362] An "emotion engine" refers to technology for recognizing a user's emotions and generating a response according to those emotions.

[1363] "Smart glasses" refers to a glasses-type device equipped with a display, camera, sensors, etc., and with the ability to display information and collect data.

[1364] "Facial expression analysis" refers to the technology of using devices such as cameras to recognize human facial expressions and estimate their emotions.

[1365] "Tone of voice analysis" refers to a technology that analyzes voice data to infer a speaker's emotions and intentions.

[1366] "Talk generation" refers to a technology that automatically creates appropriate responses and conversation content based on specific conditions and input data.

[1367] A system for implementing the present invention uses the following hardware and software.

[1368] Hardware

[1369] Smart Glasses

[1370] camera

[1371] server

[1372] software

[1373] OpenCV (face detection)

[1374] EmotionRecognizer

[1375] AITrainer (talk generation)

[1376] System Configuration

[1377] The server has the means to train the AI ​​to learn all plans and services of the company and other companies, as well as the talk of the company's top crew members. Furthermore, by combining it with an emotion engine, it can recognize the emotions of customers and generate the optimal talk according to those emotions.

[1378] Processing flow

[1379] 1. Face detection on camera

[1380] The camera on the smart glasses captures the customer's face in real time, and OpenCV is used to detect the face from the captured image.

[1381] 2. Emotion recognition

[1382] After face detection, EmotionRecognizer is used to recognize the emotion of the detected face, which identifies what emotion the customer is feeling.

[1383] 3. Token Creation

[1384] Based on the recognized emotions, AITrainer generates the optimal conversation, which has appropriate content according to the customer's emotions.

[1385] 4. Viewing chats

[1386] The generated conversations are displayed on the smart glasses' display, allowing store clerks to check the most appropriate conversations in real time and use them in conversations with customers.

[1387] Examples

[1388] For example, if a customer is looking for a new smartphone but is unsure which one to choose, they could input the following prompt into the generative AI model:

[1389] Examples of prompt statements

[1390] “What talk do you generate when a customer has a confused look on their face?”

[1391] Based on this prompt, AITrainer will generate the following talk:

[1392] Example of a generated token

[1393] "It seems like you're undecided. If you don't mind, I'd like to give you some recommendations."

[1394] In this way, AI combined with an emotion engine can generate and display optimal conversations in real time based on the customer's emotions, which is expected to improve customer satisfaction and sales efficiency.

[1395] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1396] Step 1:

[1397] A camera mounted on the smart glasses captures the customer's face in real time.

[1398] Input: Customer's facial video data

[1399] Output: The captured image frames

[1400] How it works: The camera continuously captures video and captures each frame as image data.

[1401] Step 2:

[1402] The server uses OpenCV to detect faces in the captured image frames.

[1403] Input: The captured image frame

[1404] Output: Face position information (coordinate data)

[1405] What it does: Apply OpenCV's face detection algorithm to locate the face within the image frame.

[1406] Step 3:

[1407] The server uses EmotionRecognizer to recognize the emotion of the detected face.

[1408] Input: Face position information (coordinate data) and face image

[1409] Output: Recognized emotion data (e.g., happy, sad, surprised, etc.)

[1410] What it does: A face image is fed into EmotionRecognizer, which applies an emotion recognition algorithm to classify the emotion.

[1411] Step 4:

[1412] The server uses AITrainer to generate optimal talk based on the recognized emotions.

[1413] Input: Recognized emotion data

[1414] Output: Generated chat (text data)

[1415] Specific operation: Emotion data is input into AITrainer, and appropriate speech is generated using a generative AI model.

[1416] Step 5:

[1417] The server displays the generated tokens on the display of the smart glasses.

[1418] Input: Generated token (text data)

[1419] Output: Talk displayed on the smart glasses display

[1420] Specific behavior: The generated token is sent to the smart glasses display and displayed for the user to view.

[1421] Step 6:

[1422] The user (store clerk) checks the chat displayed on the smart glasses and uses it in conversations with customers.

[1423] Input: Speech displayed on smart glasses

[1424] Output: Customer interaction

[1425] Specific operation: The store clerk reads the talk displayed on the smart glasses and interacts with the customer based on that.

[1426] Example 2

[1427] Next, a description will be given of Example 2 of the embodiment 2. In the following description, the data processing device 12 is referred to as a "server" and the smart glasses 214 are referred to as a "terminal".

[1428] In conventional sales activities, it was difficult to generate pitches to propose new plans that were differentiated from other companies' plans and services. In addition, there was a lack of ways to generate optimal pitches that matched the user's emotions, making it difficult to conduct effective sales activities. Furthermore, there was a need for efficient methods for training new employees and practicing pitches.

[1429] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and the talk of outstanding employees of the company, a means for generating talk that differentiates the company from other companies when the company issues a new plan, a means for generating counter talk when other companies issue a new plan, a means for collecting and analyzing plan information of other companies, a means for extracting differentiation points, and a means for recognizing the user's emotions and generating talk according to the emotions. This makes it possible to generate a proposal talk for a new plan that is differentiated from other companies, and to provide the optimal talk according to the user's emotions.

[1430] "Artificial intelligence" is a technology that mimics human intelligence by learning and analyzing data.

[1431] A "differentiation talk" is a proposal that emphasizes the superiority of your company's plans and services compared to those of other companies.

[1432] "Countertalk" is a proposal made to counter a new plan announced by another company.

[1433] "Plan information" is detailed information including the service content, price, conditions, etc.

[1434] "Analysis" is the process of processing collected data and understanding and classifying its contents.

[1435] A "point of differentiation" is an element that shows the characteristics or advantages of your company's plans or services that make them different from those of other companies.

[1436] "Emotion recognition" is a technology that analyzes emotions from a user's voice, facial expressions, text input, etc.

[1437] "Talk generation" is the process of automatically creating suggestions and explanations for a specific purpose.

[1438] "Employees" refers to people employed by a business.

[1439] "Training" refers to educational activities for new employees to acquire the knowledge and skills necessary for their jobs.

[1440] "Talk practice" is training for employees to effectively communicate proposals and explanations.

[1441] An "automation device" is a machine or system that performs a specific task automatically, without human assistance.

[1442] The present invention is a system for generating a proposal talk for a new plan that is differentiated from plans of other companies and providing an optimal talk according to the user's emotions. A specific embodiment of this system will be described below.

[1443] System Configuration

[1444] This system consists of three main elements: a server, a terminal, and a user. The server uses a generative AI model to collect and analyze information on plans from other companies, extracts points of differentiation, and generates a proposal for a new plan. The terminal recognizes the user's emotions and generates a proposal that corresponds to those emotions.

[1445] Hardware and Software Used

[1446] The server is a computer with high-performance computing power, and uses machine learning frameworks such as "TensorFlow (registered trademark)" and "PyTorch (registered trademark)" to execute generative AI models. It also uses web scraping technology to collect plan information from other companies. Specifically, it uses Python libraries such as "BeautifulSoup" and "Scrapy."

[1447] The terminal is a device equipped with a camera and microphone to capture the user's voice and facial expressions, and uses emotion recognition APIs such as IBM Watson and Microsoft Azure Cognitive Services.

[1448] Data processing and calculation

[1449] The server collects plan information from other companies and analyzes the content using natural language processing (NLP) technology. Based on the analysis results, the company's new plan is compared with the plans of other companies and points of differentiation are extracted. Based on the points of differentiation, the generative AI model generates a proposal speech for the new plan.

[1450] The device analyzes the user's emotions from their voice, facial expressions, text input, etc., and generates conversations according to those emotions. The emotion engine analyzes the user's emotions, and if the user is excited, it generates conversations that maintain that excitement, and if the user is anxious, it generates conversations that relieve that anxiety.

[1451] Examples

[1452] For example, when a company releases a new plan, the server collects plan information from other companies and generates differentiated talk using a generative AI model. When a user inquires about the new plan, the device analyzes the user's emotions and generates the optimal talk based on those emotions.

[1453] An example of a specific prompt is as follows:

[1454] Generate differentiated talk for your new plan based on other companies' plan information. For example, if the other company's plan is "1000 yen per month with unlimited calls," generate a differentiated talk for your new plan. Also, generate talk according to the user's emotions. Consider both scenarios when the user is excited and when he or she is anxious.

[1455] In this way, a system is realized in which the server, terminal, and user work together to generate proposal talks for new plans that are differentiated from those of other companies, and provide optimal talks that correspond to the user's emotions.

[1456] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1457] Step 1: Gather information about other companies' plans

[1458] The server collects plan information from other companies. It receives publicly available information on the Internet and database URLs as input. Specifically, the server uses web scraping technology to obtain plan information from other companies' official websites. It uses Python libraries such as "BeautifulSoup" and "Scrapy" to analyze pages containing plan information. It obtains the collected plan information as text data as output.

[1459] Step 2: Analyze other companies' plans

[1460] The server analyzes the collected plan information of other companies. As input, it receives the text data collected in step 1. Specifically, it uses natural language processing (NLP) technology to extract the plan contents and features. It analyzes the text data using "TensorFlow (registered trademark)" and "PyTorch (registered trademark)" to extract the plan price and service details. It obtains the analysis results as output.

[1461] Step 3: Extracting points of differentiation

[1462] The server compares the company's new plan with other companies' plans based on the analysis results and extracts points of differentiation. As input, it receives the analysis results obtained in step 2 and the company's plan information. Specifically, it compares the plan information of other companies with the company's plan information and lists the differences in price and service content. As output, it obtains points of differentiation.

[1463] Step 4: Generate a new plan proposal token

[1464] The server generates a proposal speech for the new plan using the generative AI model based on the differentiation points. It receives the differentiation points obtained in step 3 as input. In concrete terms, it inputs a prompt sentence into the generative AI model to generate a proposal speech for the new plan. As an example of a prompt sentence, it inputs, "If the plan of another company is 1,000 yen per month with unlimited calling charges, please generate a proposal speech for a differentiated new plan." As output, it obtains the generated proposal speech.

[1465] Step 5: Recognizing user emotions

[1466] The device recognizes the user's emotions. It receives the user's voice, facial expressions, and text input as input. Specifically, it sends data captured by the camera and microphone to an emotion recognition API. It then uses emotion recognition APIs such as "IBM Watson" and "Microsoft Azure Cognitive Services" to analyze the user's emotions. It obtains the analysis results as output.

[1467] Step 6: Generate speech based on emotions

[1468] The server generates talk according to the user's emotions based on the analysis results of the emotion engine. As input, it receives the emotion analysis results obtained in step 5 and the suggested talk generated in step 4. In concrete terms, it inputs a prompt sentence into the generative AI model and generates talk according to the emotion. As an example of a prompt sentence, it inputs "If the user is excited, please generate talk that will maintain that excitement." As output, it obtains the optimal talk according to the emotion.

[1469] (Application example 2)

[1470] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server", and the smart glasses 214 will be referred to as a "terminal".

[1471] In conventional sales activities, it was difficult to generate a sales pitch that differentiated the company from its competitors, and it was not possible to provide the optimal sales pitch according to the user's emotions. In particular, online shopping sites are required to provide differentiated sales pitches in real time while users are browsing products, but there was no effective way to achieve this.

[1472] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1473] In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and the talk of the company's excellent employees, a means for generating talk that differentiates the company from other companies when the company releases a new plan, a means for generating counter talk when other companies release a new plan, a means for recognizing the user's emotions and generating talk according to the emotions, a means for collecting information on other companies and learning their points of differentiation, and a means for generating optimal talk according to the user's emotions. This makes it possible to provide optimal differentiated talk according to the user's emotions in real time.

[1474] "All plans and services of our company and competitors" refers to all rate plans and service contents offered by our company and competitors.

[1475] "Top employee talk" refers to effective sales talk used by employees who are particularly successful in sales activities.

[1476] "Artificial intelligence" refers to systems that mimic human intelligence using technologies such as machine learning and natural language processing.

[1477] "Differentiation talk" refers to sales talk that emphasizes the superiority of one's company's plans or services compared to those of other companies.

[1478] "Countertalk" refers to sales pitches made to counter an announcement of a new plan or service by another company.

[1479] "Means of recognizing the user's emotions" refers to technology that uses sensors such as cameras and microphones to analyze the user's emotions from their facial expressions and tone of voice.

[1480] "Means for generating a talk according to emotions" refers to a technology for generating an optimal sales talk based on the recognized emotions of a user.

[1481] "Means of collecting information about other companies" refers to technology for collecting information about competitors' plans and services from the Internet, etc.

[1482] "Means of learning points of differentiation" refers to technology that learns in what ways a company's plans and services are superior based on collected information about other companies.

[1483] "Providing in real time" refers to generating and providing an appropriate sales pitch at the exact moment that the user is viewing a product.

[1484] As an embodiment of the present invention, the following system is constructed.

[1485] System Configuration

[1486] The system consists of a server, a user terminal (smartphone), and various software components. The server hosts an artificial intelligence (AI) model and processes data from the user terminal. The user terminal recognizes the user's emotions using a camera and microphone and transmits them to the server.

[1487] Hardware and software used

[1488] Hardware: Smartphone (camera, microphone)

[1489] Software: TensorFlow (registered trademark) (AI model), OpenCV (emotion recognition), Flask (backend), React Native (frontend)

[1490] Program Processing

[1491] 1. Data collection:

[1492] The server scrapes product information from other companies' online shopping sites and stores it in a database. This allows us to collect information about other companies' plans and services. The libraries used are BeautifulSoup and Requests.

[1493] 2. Training the AI ​​model:

[1494] The server learns the points of differentiation between the company's products and those of other companies based on the collected data. The library used is TensorFlow (registered trademark).

[1495] 3. Emotion recognition:

[1496] The user device uses a camera and microphone to analyze the user's facial expressions and tone of voice in real time. The libraries used are OpenCV and Librosa.

[1497] 4. Token creation:

[1498] The server generates differentiated speech based on the user's emotions. The library used is GPT-3 (registered trademark) (OpenAI (registered trademark)).

[1499] 5. Frontend display:

[1500] The user's device uses React Native to display the generated token to the user.

[1501] Examples

[1502] For example, if another company's plan is "unlimited calling charges for 1,000 yen per month," when the user is excited, a conversation might be generated that says, "unlimited calling charges and unlimited data charges for 500 yen per month."

[1503] Examples of prompt statements

[1504] If another company's plan is "unlimited calls for 1,000 yen per month," generate a conversation while your users are excited that says "unlimited calls and unlimited data for 500 yen per month."

[1505] In this way, it is possible to provide optimal differentiated talk in real time according to the user's emotions.

[1506] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1507] Step 1:

[1508] The server scrapes product information from other companies' mail order sites and stores it in a database. Specifically, it uses BeautifulSoup and Requests to obtain information such as product names, prices, and plan details from other companies' web pages. The input is the URL of the other companies' web pages, and the output is the storage of the obtained product information in the database.

[1509] Step 2:

[1510] The server learns the points of differentiation between the company's products and those of other companies based on the collected data. Specifically, it uses TensorFlow (registered trademark) to input the collected data into an AI model and extract the points of differentiation. The input is product information stored in a database, and the output is an AI model that has learned the points of differentiation.

[1511] Step 3:

[1512] The user device uses a camera and microphone to analyze the user's facial expressions and tone of voice in real time. Specifically, it uses OpenCV and Librosa to recognize the user's emotions. The input is the video and audio data captured by the camera and microphone, and the output is the recognized user's emotion data.

[1513] Step 4:

[1514] The server generates differentiated speech based on the recognized user emotion. Specifically, it uses GPT-3 (registered trademark) (OpenAI (registered trademark)) to input the user's emotion data and differentiation points as prompts and generate optimal speech. The input is the user's emotion data and differentiation points, and the output is the generated speech.

[1515] Step 5:

[1516] The user device uses React Native to display the generated token to the user. Specifically, the generated token is displayed on the screen and provided to the user. The input is the generated token, and the output is the token displayed to the user.

[1517] Example 3

[1518] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[1519] In conventional sales activities, it was difficult to generate counter talk to quickly and effectively counter new plans from other companies, or to create new plan / service proposals. In addition, when training new employees, there was often a lack of practice in effective talk. Furthermore, there were limited means to improve the efficiency of customer acquisition. To solve these problems, a system was needed that could quickly collect information on new plans from other companies, and use that information to generate effective counter talk and new plan / service proposals.

[1520] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1521] In this invention, the server includes a means for collecting new plan information from other companies, a means for creating a prompt sentence to be input to the generation AI model based on the collected information, and a means for generating counter talk using the generation AI model. This makes it possible to generate counter talk that quickly and effectively counters new plans from other companies. In addition, by including a means for creating new plan / service proposals that exploit the weaknesses of other companies' plans and services, a means for using the plan as a talking wall material during training for new employees, a means for using an emotion engine that recognizes the user's emotions and generates talk according to the emotions, and a means for sending the generated talk to a robot and for the robot to automatically conduct sales activities to customers, it becomes possible to significantly improve the efficiency and effectiveness of sales activities.

[1522] "Means of collecting information on new plans from other companies" refers to means of collecting information on new plans and services announced by other companies using web scraping tools or APIs.

[1523] The "means for creating prompt sentences to be input into the generative AI model" refers to a means for creating instruction sentences for causing the generative AI model to generate appropriate countertalk or new plan / service proposals based on collected information on new plans from other companies.

[1524] "Means for generating countertalk using a generative AI model" refers to a means for inputting a prompt sentence into a generative AI model to generate countertalk to counter a new plan of another company.

[1525] "Means for creating new plan / service proposals that exploit weaknesses in the plans and services of other companies" refers to means for generating new plan / service proposals that exploit weaknesses using a generative AI model based on detailed information about the plans and services of other companies.

[1526] "A means of using the talk generated by the generative AI model as material for bouncing talks off of during training for new employees" refers to a means of using the talk generated by the generative AI model during training for new employees to practice and simulate talks.

[1527] "Means of using an emotion engine that recognizes the user's emotions and generates talk in accordance with those emotions" refers to a means of analyzing the user's emotions with an emotion engine and, based on the results, a generative AI model generating optimal talk in accordance with the user's emotions.

[1528] "A means of sending the generated talk to a robot, which then automatically carries out sales activities toward customers" refers to a means of sending the talk generated by a generative AI model to a robot, which then automatically carries out sales activities toward customers based on that talk.

[1529] This invention is a system that generates counter talk to new plans when other companies announce new plans. The system collects information on new plans from other companies and uses a generative AI model to generate counter talk and new plan / service proposals. It can also be used to improve the efficiency of training new employees and customer acquisition.

[1530] Hardware and Software Used

[1531] Hardware

[1532] Server: Use a server with a high-performance processor and a large amount of memory. For example, use a virtual server provided by a cloud service provider (e.g., Amazon EC2, Google Cloud Compute Engine).

[1533] Device: Computers and tablets used by new employees.

[1534] Robots: Robots that automatically conduct sales activities with customers (e.g. Pepper, NAO).

[1535] software

[1536] Generative AI model: We use GPT-4 (registered trademark) from OpenAI (registered trademark) as the generative AI model.

[1537] Web scraping tools: We use BeautifulSoup and Scrapy to gather information on new plans from other companies.

[1538] Emotion Engine: Uses Affectiva and IBM Watson Tone Analyzer to recognize user emotions.

[1539] Data processing and calculation

[1540] Collecting information on new plans from other companies

[1541] The server uses web scraping tools (e.g. BeautifulSoup, Scrapy) to collect new plan information from other companies' official websites and news sites. The collected information is stored in a database.

[1542] Creating a prompt statement

[1543] The server creates a prompt sentence to be input to the generative AI model based on the collected information on new plans from other companies. As a specific example, the server generates the following prompt sentence:

[1544] "Another company has announced a plan offering unlimited calls and data for 500 yen per month. Please generate a counter-talk to this plan."

[1545] Generating counter talk

[1546] The server inputs the prompt sentence into a generative AI model (e.g., GPT-4 (registered trademark) of OpenAI (registered trademark)) to generate countertalk. The generated countertalk is stored in a database.

[1547] Creating new plans and service proposals

[1548] Based on detailed information about other companies' plans and services, the server inputs prompt sentences like the following into the generative AI model to generate new plan and service proposals.

[1549] "Generate new plan and service proposals that exploit the weaknesses of other companies' plans and services."

[1550] Training new employees

[1551] The user (new employee) uses the talk generated by the generative AI model as material for bouncing ideas off the wall. The server generates a prompt sentence like the one below and sends it to the generative AI model.

[1552] "Generate counter talks to other companies' plans for onboarding new employees."

[1553] Robot customer acquisition

[1554] The server sends the chat generated by the generative AI model to the robot. The robot automatically uses the chat to conduct sales activities with customers. As a concrete example, the following chat is generated:

[1555] "Hello, let me introduce you to our new plan. For just 500 yen per month you get unlimited calls, unlimited data, and 24 / 7 support."

[1556] Using the Emotion Engine

[1557] The server recognizes the user's emotions using an emotion engine (e.g., Affectiva, IBM Watson Tone Analyzer). Based on the analysis results, the following prompt sentences are input into the generative AI model to generate a conversation according to the user's emotions.

[1558] "If a user is unhappy with another company's new plan, generate a chat that addresses that frustration."

[1559] In this way, the system can generate counter talks and new plan / service proposals to compete with new plans from other companies, and provide optimal talks according to the user's emotions. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1560] Step 1:

[1561] The server collects new plan information from other companies. Specifically, it uses web scraping tools (e.g. BeautifulSoup, Scrapy) to obtain new plan information from other companies' official websites and news sites. As input, it uses the URLs and API endpoints of other companies' websites, and as output, it stores the new plan information in a database. The server periodically checks these sites and updates the database if there is new information.

[1562] Step 2:

[1563] The server creates a prompt sentence to be input to the generative AI model based on the collected information on new plans from other companies. It uses the new plan information stored in the database as input, and generates a prompt sentence to send to the generative AI model as output. In concrete terms, the server creates a prompt sentence such as, "Another company has announced a plan with unlimited calls and unlimited data communication for 500 yen per month. Please generate a countertalk to counter this."

[1564] Step 3:

[1565] The server inputs the prompt sentence into a generative AI model (e.g., GPT-4 (registered trademark) by OpenAI (registered trademark)) and generates countertalk. As input, it uses the prompt sentence created in step 2, and as output, it stores the generated countertalk in a database. In concrete terms, the server sends the prompt sentence to the generative AI model and receives the countertalk returned from the model.

[1566] Step 4:

[1567] The server creates a prompt sentence to be input into the generative AI model based on detailed information about other companies' plans and services, and generates new plan / service proposals. Detailed information about other companies' plans and services stored in the database is used as input, and a prompt sentence is created to generate new plan / service proposals as output. Specifically, the server generates a prompt sentence saying, "Please generate new plan / service proposals that exploit the weaknesses of other companies' plans and services."

[1568] Step 5:

[1569] The user (new employee) uses the talk generated by the generative AI model as material for bouncing ideas off the wall. The generated talk is used as input, and is provided as output as training material for the new employee. In concrete terms, the server generates a prompt message saying, "Please generate a talk that competes with other companies' plans for the training of new employees," and sends it to the generative AI model. The generated talk is provided as training material for the new employee.

[1570] Step 6:

[1571] The server sends the generated talk to the robot. The generated talk is used as input, and the robot automatically conducts sales activities to customers as output. In concrete terms, the server sends the generated talk to the robot's control system, and the robot automatically conducts sales activities to customers based on the talk.

[1572] Step 7:

[1573] The server recognizes the user's emotions using an emotion engine (e.g. Affectiva, IBM Watson Tone Analyzer). The user's voice and text data are used as input, and the user's emotional state is analyzed as output. Based on the analysis results, a prompt sentence is input into the generative AI model, and a conversation corresponding to the user's emotions is generated. In concrete terms, the server inputs the user's voice and text data into the emotion engine, and analyzes the user's emotional state. Based on the analysis results, a prompt sentence is input into the generative AI model, such as, "If the user is dissatisfied with another company's new plan, please generate a conversation that will resolve that dissatisfaction," and the generated conversation is provided to the user.

[1574] (Application example 3)

[1575] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server", and the smart glasses 214 will be referred to as a "terminal".

[1576] In conventional sales activities, when other companies announced new plans, it was difficult to generate a counter pitch quickly and effectively. In addition, it was not possible to provide the optimal pitch according to the customer's emotions, making it difficult to improve customer satisfaction and acquire customers efficiently. Furthermore, there was an issue of a lack of practical pitches during training for new crew members.

[1577] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1578] In this invention, the server includes a means for making the AI ​​learn all plans and services of the company and other companies, and the talk of the company's excellent crew, a means for generating talk that differentiates the company from other companies when the company issues a new plan, a means for generating counter talk when other companies issue a new plan, a means for combining an emotion engine that recognizes customer emotions and generates talk according to the emotions, and a means for displaying talk on a smart device. This allows the company to provide optimal talk according to customer emotions in real time, improving customer satisfaction and efficiently acquiring customers.

[1579] "All plans and services of our company and competitors" refers to all rate plans and service contents offered by our company and competitors.

[1580] "Excellent crew talk" refers to effective sales talk used by crew members who have particularly excellent sales results within a company.

[1581] "Means of training an AI" refers to the methods or processes for using artificial intelligence technology to train specific data or information.

[1582] "Means for generating differentiating talk" refers to a method for automatically generating talk to emphasize the superiority of a company compared to the plans or services of other companies.

[1583] "Means for generating countertalk" refers to a method for automatically generating countertalk when another company announces a new plan.

[1584] An "emotion engine" refers to technology that recognizes a user's emotions and generates an optimal response based on those emotions.

[1585] "Means for displaying talk on a smart device" refers to a method for displaying the generated talk on a device such as smart glasses or a smartphone.

[1586] As an embodiment of the present invention, a system using smart glasses in a brick-and-mortar store will be described as an example.

[1587] First, the server trains the AI ​​to learn all plans and services of the company and other companies, as well as the talk of the company's top crew members. This gives the AI ​​the ability to understand the characteristics of each plan and service and generate appropriate talk.

[1588] Next, when the server launches a new plan, it creates a differentiation pitch from other companies. For example, if another company launches a plan that offers "unlimited calls and unlimited data for 500 yen per month," the server creates a counter-pitch that says, "unlimited calls and unlimited data for 500 yen per month, plus 24 / 7 support."

[1589] In addition, the server combines an emotion engine that recognizes the customer's emotions and generates speech according to those emotions. The emotion engine analyzes the customer's facial expressions and tone of voice captured by the camera, and generates speech to calm the customer's excitement if the customer is excited, and generates speech to resolve the customer's dissatisfaction if the customer is dissatisfied.

[1590] Smart glasses are used as a means of displaying talk on smart devices. Salespeople wear the smart glasses and see counter talk generated by AI and talk according to emotions in real time while interacting with customers. This allows salespeople to instantly provide the most appropriate talk according to the customer's questions and their reactions to other companies' plans.

[1591] For example, if a customer presents a plan that offers "unlimited calls and data for 500 yen per month," the AI ​​generates a countertalk of "unlimited calls and data for 500 yen per month, plus 24 / 7 support." Furthermore, if the customer is excited about this plan, the emotion engine generates a countertalk that calms that excitement.

[1592] Examples of prompts to input to a generative AI model include:

[1593] "Another company has announced a new plan: unlimited calls and unlimited data for 500 yen per month. Please generate a counter-talk to this."

[1594] "Your customers are excited. Generate conversations that reflect this emotion: 500 yen per month for unlimited calls, unlimited data, and 24 / 7 support."

[1595] In this way, it is possible to provide the most appropriate conversation in real time based on the customer's emotions, improving customer satisfaction and efficiently acquiring new customers.

[1596] The flow of the specific process in the application example 3 will be described with reference to FIG.

[1597] Step 1:

[1598] The server trains the AI ​​to learn all plans and services from the company and other companies, as well as the conversations of the company's top crew members. Specifically, it collects plan and service information and conversation data from the database and inputs them into the AI ​​model. The AI ​​model learns this data and understands the characteristics of each plan and service. The input is plan and service information and conversation data, and the output is the trained AI model.

[1599] Step 2:

[1600] When a company issues a new plan, the server generates a differentiation token that differentiates the company from other companies. Specifically, when a user inputs new plan information, the AI ​​model compares it with other companies' plans and extracts points of differentiation. Based on this, a differentiation token is generated. The input is the new plan information, and the output is the differentiation token.

[1601] Step 3:

[1602] The server generates countertalk when another company releases a new plan. Specifically, it inputs the new plan information of the other company into the AI ​​model and generates countertalk. The input is the new plan information of the other company, and the output is the countertalk.

[1603] Step 4:

[1604] The server combines an emotion engine that recognizes the customer's emotions and generates speech corresponding to those emotions. Specifically, the customer's facial expressions and tone of voice captured by the camera are input into an emotion recognition model to analyze the customer's emotions. Based on the analysis results, the AI ​​model generates speech corresponding to the emotion. The input is the customer's facial expressions and tone of voice, and the output is speech corresponding to the emotion.

[1605] Step 5:

[1606] The terminal displays the talk on the smart device. Specifically, the terminal displays the generated talk on the display of the smart glasses. The input is the generated talk, and the output is the talk displayed on the smart glasses.

[1607] Step 6:

[1608] The user interacts with the customer based on the talk displayed through the smart glasses. Specifically, the user provides the customer with the most appropriate talk while referring to the talk displayed on the smart glasses. The input is the talk displayed on the smart glasses, and the output is the conversation with the customer.

[1609] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires a voice indicating a user input for the result of the specific processing. The control unit 46A transmits the voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1610] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search<URL: https: / / OpenAI.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data according to the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1611] Another example of generative AI is Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[1612] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1613] [Third embodiment]

[1614] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1615] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1616] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a wide area network (WAN) and / or a local area network (LAN).

[1617] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1618] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs the voice according to instructions from the processor 46.

[1619] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[1620] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[1621] Fig. 6 shows an example of main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1622] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32, and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1623] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1624] In the headset type terminal 314, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50, and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1625] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1626] "Example 1"

[1627] The present invention has a means for making AI learn all plans and services of the company and other companies, and the talk of the company's excellent crew. Specifically, information on plans and services of the company and other companies, and the contents of the talk of the company's excellent crew are stored in a database, and the AI ​​learns this information.

[1628] "Example 2"

[1629] Furthermore, when a company releases a new plan, it has the means to generate a differentiating message from other companies. Specifically, based on the information the AI ​​has learned, it generates a proposal message for a new plan that is differentiated from other companies' plans and services. For example, if another company's plan is "unlimited calling charges for 1,000 yen per month," the AI ​​generates a differentiating message that says "unlimited calling charges and unlimited data communication for 500 yen per month."

[1630] "Example 3"

[1631] In addition, it has a means to generate counter talk when other companies release new plans. Specifically, when another company releases a new plan, the AI ​​learns information about that plan and generates counter talk to counter it. For example, if another company releases a plan that offers "unlimited calls and unlimited data for 500 yen per month," the AI ​​generates the counter talk "unlimited calls and unlimited data for 500 yen per month, plus 24 / 7 support."

[1632] "Example 4"

[1633] In addition, it has the means to create new plans and service proposals that exploit the weaknesses of other companies' plans and services, and to use them as a conversation wall during training for new crew members. Specifically, based on the information that AI has learned, it creates new plans and service proposals that exploit the weaknesses of other companies' plans and services. In addition, during training for new crew members,

[1634] The talk generated by AI is used as material for bouncing ideas off each other.

[1635] "Example 5"

[1636] Finally, in the future, we will have the means to improve the efficiency of customer acquisition by transferring the acquisition to robots. Specifically, the robot will use the chat generated by AI to improve the efficiency of customer acquisition. For example, the robot will automatically conduct sales activities to customers based on the chat generated by AI.

[1637] The process flow of each embodiment will be described below.

[1638] "Example 1"

[1639] Step 1: Accumulate a database of all plans and services from your company and other companies, as well as information on the talks of your company's top crew members.

[1640] Step 2: The AI ​​learns the information from the database.

[1641] "Example 2"

[1642] Step 1: Based on the information the AI ​​has learned, it generates a proposal speech for a new plan that is differentiated from other companies' plans and services.

[1643] Step 2: Provide the generated chat to the sales team and use it to propose new plans.

[1644] "Example 3"

[1645] Step 1: When another company releases a new plan, the AI ​​learns the information about that plan.

[1646] Step 2: Based on the information that the AI ​​has learned, it generates a counter talk to counter the new plan. Step 3: The generated counter talk is provided to the sales team and used in sales activities to counter the new plan of other companies.

[1647] "Example 4"

[1648] Step 1: Based on the information the AI ​​has learned, it creates new plan and service proposals that exploit the weaknesses of other companies' plans and services.

[1649] Step 2: When training new crew members, use the AI-generated talk as a training material.

[1650] "Example 5"

[1651] Step 1: The robot uses the AI-generated chat to improve the efficiency of customer acquisition. Step 2: Based on the AI-generated chat, the robot automatically conducts sales activities to customers.

[1652] Example 1

[1653] Next, a description will be given of Example 1 of the embodiment 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal".

[1654] With the previous system, it was difficult to effectively utilize plan and service information from the company and other companies, as well as the dialogue content of excellent employees, making it difficult to provide quick and appropriate answers to user inquiries. In addition, efficient methods were required for creating new plan and service proposals and training new employees. Furthermore, there was a lack of automated means to improve the efficiency of customer acquisition.

[1655] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1656] In this invention, the server includes a means for collecting plan and service information of the company and other companies, and the contents of dialogue of the company's excellent employees, and storing them in a database, a means for preprocessing the stored data and learning them using a generative AI model, and a means for analyzing inquiries from users and generating optimal answers using the generative AI model. This makes it possible to provide prompt and appropriate answers to inquiries from users. It also provides an efficient method for creating new plan and service proposals and training new employees. Furthermore, it is possible to realize an automation means for improving the efficiency of customer acquisition.

[1657] "Plan and service information" is detailed data on the various services and pricing plans offered by companies.

[1658] "Conversations between excellent employees" are records of actual conversations conducted by employees who were highly evaluated for their customer service.

[1659] A "database" is an information accumulation system that systematically stores collected data and enables efficient search and use.

[1660] "Preprocessing" refers to processes such as data cleaning, normalization, and tokenization to prepare collected data in a format suitable for an AI model.

[1661] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate answers to user inquiries.

[1662] A "user query" is a question or information request made by a user to the system.

[1663] "Answer generation" is the process of creating an answer based on appropriate information to a user's inquiry.

[1664] A "terminal" is a device through which a user accesses and queries the system.

[1665] "Dialogue simulation" is a training method that allows new employees to gain simulated experience interacting with real customers.

[1666] "Customer acquisition automation" is a technology that automates the process of efficiently acquiring customers.

[1667] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[1668] The present invention is a system for providing prompt and appropriate responses to inquiries from users. A specific embodiment of this system will be described below.

[1669] 1. Program Generation

[1670] The server generates a program that collects the plan and service information of the company and other companies provided by the user, as well as the conversations of the company's outstanding employees, and stores them in a database. This program is developed using programming languages ​​such as Python and Java (registered trademark).

[1671] 2. Data collection and storage

[1672] The server stores the collected data in a database. A relational database such as MySQL (registered trademark) or PostgreSQL (registered trademark) is used as the database. Specifically, the following data is collected:

[1673] Detailed information about your plans and services (e.g., pricing, offerings, terms and conditions)

[1674] Detailed information about other companies' plans and services (e.g., fees, services provided, contract terms)

[1675] Conversations by your company's outstanding employees (e.g. customer service scripts, success stories)

[1676] 3. Data Preprocessing

[1677] The server preprocesses the data stored in the database, specifically cleaning the data (filling in missing values ​​and removing outliers), normalizing it, tokenizing it, etc. This prepares the data in a format suitable for the generative AI model.

[1678] 4. Training the model

[1679] The server trains a generative AI model using the preprocessed data. The AI ​​model used is a natural language processing model such as GPT-3 (registered trademark) or BERT. The learning process uses machine learning frameworks such as TensorFlow (registered trademark) or PyTorch (registered trademark).

[1680] 5. Parsing the query and generating an answer

[1681] When a user makes an inquiry through a terminal, the server analyzes the content of the inquiry. It uses natural language processing technology to understand the user's intent and extract the necessary information. The server uses a generative AI model based on the analysis results to generate a response. Specifically, it combines the optimal plan and service information and dialogue content for the user's inquiry to create a response in a natural dialogue format.

[1682] 6. Providing a response

[1683] The server sends the generated answer to the user's terminal, through which the user can check the answer and make additional inquiries if necessary.

[1684] Examples

[1685] For example, if a user asks "What is the difference between my company's plan A and another company's plan B?" the server would process it as follows:

[1686] 1. Analyze the user's query and retrieve information for Plan A and Plan B from the database.

[1687] 2. Compare the information obtained and clarify the differences.

[1688] 3. Generate answers in a natural, conversational format based on the conversations of your top employees.

[1689] Examples of prompt statements

[1690] Examples of prompts to be fed to a generative AI model include:

[1691] User: "What is the difference between my Plan A and the other company's Plan B?"

[1692] AI model: "Our company's Plan A costs 5,000 yen per month and has unlimited data. The other company's Plan B costs 4,500 yen per month and has a data limit of 10GB. In addition, Plan A includes free calls, but Plan B does not."

[1693] In this way, the server realizes a system that provides appropriate answers to user inquiries.

[1694] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1695] Step 1: Data collection

[1696] The server collects plan and service information from the company and other companies, as well as the content of conversations between the company's outstanding employees. Specifically, it uses APIs and scraping tools to obtain the necessary data from websites and internal systems. The inputs include various API endpoints and webpage URLs, and the output is the acquired plan and service information and conversation content data.

[1697] Step 2: Data accumulation

[1698] The server stores the collected data in a database. MySQL (registered trademark) or PostgreSQL (registered trademark) is used as the database. Specifically, it executes SQL queries to insert the acquired data into appropriate tables. The inputs include the collected plan and service information and conversation content data, and the output is the data stored in the database.

[1699] Step 3: Data Preprocessing

[1700] The server preprocesses the data stored in the database, specifically cleaning the data (filling in missing values ​​and removing outliers), normalizing it, tokenizing it, etc. The input is the raw data stored in the database, and the output is the preprocessed, clean data.

[1701] Step 4: Model training

[1702] The server trains a generative AI model using the preprocessed data. The AI ​​models used are GPT-3 (registered trademark) and BERT, and training is performed using TensorFlow (registered trademark) and PyTorch (registered trademark). Specifically, the data is divided into a training set and a test set, and the accuracy of the model is evaluated. The input is the preprocessed data, and the output is a trained AI model.

[1703] Step 5: Query Analysis

[1704] When a user makes a query through a terminal, the server analyzes the query. It uses natural language processing technology to understand the user's intent and extract the necessary information. The input is the query text from the user, and the output is the analyzed query.

[1705] Step 6: Answer Generation

[1706] The server generates an answer using a generative AI model based on the analysis results. Specifically, it creates an answer in a natural dialogue format by combining the optimal plan and service information and dialogue content for the user's inquiry. The input is the analyzed inquiry content and the trained AI model, and the output is the generated answer text.

[1707] Step 7: Provide your answer

[1708] The server sends the generated answer to the user's terminal, where the user can check the answer and make additional inquiries if necessary. The input is the generated answer text, and the output is the answer displayed on the user's terminal.

[1709] (Application example 1)

[1710] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server", and the headset type terminal 314 will be referred to as a "terminal".

[1711] With the conventional system, it was difficult to effectively utilize plan and service information from our company and other companies to make optimal proposals to users. In addition, there was a lack of a way to provide optimal advice to users through chat using AI that learned the conversation content of excellent crew members. This resulted in a lack of improvement in the user's purchasing experience, and there were concerns about a decline in competitiveness.

[1712] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1713] In this invention, the server includes a means for making an AI learn all plans and services of the company and other companies, and the talk of the company's top crew, a means for generating talk that differentiates the company from other companies when the company releases a new plan, a means for generating counter talk when other companies release a new plan, a means for proposing the best plan or service for a product that the user is considering purchasing, a means for providing the user with the best option based on the company's and other companies' plan and service information, and a means for the AI ​​that has learned the talk content of the company's top crew to provide the best advice in a chat with the user. This makes it possible to propose the best plan or service to the user and improve the purchasing experience.

[1714] "Plan / Service" refers to the content of the products and services offered by the company and their terms of use.

[1715] "Excellent crew" refers to employees within a company who have particularly outstanding performance and skills.

[1716] "Talk" refers to the way of speaking and the content of explanations used in customer interactions and sales activities.

[1717] "AI" is an abbreviation for artificial intelligence and refers to systems that mimic human intelligence using technologies such as machine learning and natural language processing.

[1718] "Differentiation talk" refers to the way of speaking and the content of explanations that emphasize the superiority of a company's plans and services compared to those of other companies.

[1719] "Countertalk" refers to the way of speaking or the explanation given when another company announces a new plan or service.

[1720] "User" refers to a customer or consumer who uses the company's plans or services.

[1721] "Chat" refers to a means of communicating in real time using text messages.

[1722] "Buying experience" refers to the overall experience and satisfaction a user feels when purchasing a product or service.

[1723] "Choices" refers to the selection of plans or services available to a User.

[1724] The system for implementing this invention has the following configuration. The server accumulates information on plans and services from the company and other companies, as well as the contents of the conversations of the company's top crew members in a database, and has the AI ​​learn this information. When a user tries to purchase a product using their smartphone, the server receives the user's input and suggests the most suitable plan and service.

[1725] Hardware and Software Used

[1726] Hardware: Smartphones, servers

[1727] Software: Python, OpenAI® API

[1728] Data processing and calculation

[1729] The server processes the data in the following steps:

[1730] 1. Building the database:

[1731] The server stores information on plans and services from the company and other companies, as well as the contents of the conversations of the company's outstanding crew members in a database, allowing the AI ​​to learn from this information.

[1732] 2. Receiving user input:

[1733] The user inputs information about the product he or she is considering purchasing using a smartphone. For example, the user might input, "I want to buy a new smartphone."

[1734] 3. Generate prompt:

[1735] The server generates prompts based on user input, including information about plans and services from the company and other companies, and the content of talks by outstanding crew members.

[1736] 4. AI-generated responses:

[1737] The server sends the generated prompts to the OpenAI® API and receives an AI-generated response, which includes recommendations for the best plan or service for the user.

[1738] 5. Display response to user:

[1739] The server then displays the response on the user's smartphone, allowing the user to select the most suitable plan or service.

[1740] Examples

[1741] If the user types "I want to buy a new smartphone", the server generates a prompt like this:

[1742] Example prompt:

[1743] What people are considering buying: A new smartphone

[1744] Please suggest the best plan and service for you.

[1745] Company Plans & Services: Plan A, Plan B, Plan C

[1746] Other companies' plans and services: Plan X, Plan Y, Plan Z

[1747] Topic of the excellent crew: This plan is very cost-effective. This service is especially recommended for beginners. This product is currently on sale and is a great deal.

[1748] Based on these prompts, the AI ​​can generate responses suggesting the best plan or service for the user, for example, "Plan A is currently on sale and is a good deal. Plan B is recommended for beginners."

[1749] This way, users can choose the best plan or service for them, improving their purchasing experience.

[1750] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1751] Step 1:

[1752] The server stores the plan and service information of the company and other companies, as well as the contents of the conversations of the company's outstanding crew members in a database. This allows the AI ​​to learn from this information. The input is the plan and service information and the contents of the conversations, and the output is a database in which this information is stored. In concrete terms, the server uses a database management system (DBMS) to organize the information and save it in an appropriate format.

[1753] Step 2:

[1754] A user uses a smartphone to input information about a product he or she is considering purchasing. For example, the input may be, "I want to buy a new smartphone." The input includes information about the product the user is considering, and the output is that information is sent to a server. Specifically, the user inputs text through a smartphone application, and the data is sent to the server.

[1755] Step 3:

[1756] The server generates a prompt based on the user's input. The prompt includes the company's and other companies' plan and service information, and the talk content of outstanding crew members. The input is the user's input information, as well as the plan and service information and talk content stored in the database, and the generated prompt is obtained as the output. In terms of specific operation, the prompt is constructed using a Python script.

[1757] Step 4:

[1758] The server sends the generated prompt to the OpenAI (registered trademark) API and obtains a response from the AI. The generated prompt is the input, and the AI's response is the output. The specific operation is to call the OpenAI (registered trademark) API, send a prompt, and receive a response.

[1759] Step 5:

[1760] The server displays the obtained response on the user's smartphone. This allows the user to select the most suitable plan and service. The input is the response from the AI, and the output is the information displayed on the user's smartphone. The specific operation is to update the UI (user interface) to display the response through the smartphone application.

[1761] In this way, users can select the most suitable plan or service, improving their purchasing experience.

[1762] Example 2

[1763] Next, a second embodiment of the second form will be described. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal".

[1764] In the conventional system, there was a problem that it took a lot of time and effort to generate counter talks to differentiate plans and services between one's own and those of other companies. In addition, it was difficult to quickly generate counter talks when other companies announced new plans, making it difficult to maintain competitiveness. Furthermore, there was an issue of a lack of practice in effective talks in training new employees.

[1765] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1766] In this invention, the server includes a means for making an artificial intelligence learn all plans and services of the company and other companies, and talk of the company's outstanding employees, a means for collecting plan and service information of other companies, a means for preprocessing the collected data, a means for inputting the preprocessed data into a generating AI model, a means for generating a differentiation talk from other companies using the generating AI model, a means for outputting the generated differentiation talk, and a means for generating a counter talk when another company issues a new plan. This makes it possible to generate differentiation talk and counter talk quickly and efficiently.

[1767] "All plans and services of our company and competitors" refers to all pricing plans and service contents offered by our company and competitors.

[1768] "Excellent employee talk" refers to the effective speaking techniques and sales pitches that a company's top employees use when dealing with customers and in sales activities.

[1769] "Artificial intelligence" refers to computer systems that mimic human intelligence using techniques such as machine learning and natural language processing.

[1770] "Means of collecting information on the plans and services of other companies" refers to methods of automatically obtaining competitors' pricing plans and service details using web scraping tools, APIs, etc.

[1771] "Means for preprocessing collected data" refers to methods for analyzing acquired data and processing it to delete unnecessary information or standardize the format.

[1772] "Means of inputting preprocessed data into a generative AI model" refers to a method of providing preprocessed data in an appropriate format to a generative AI model.

[1773] A "generative AI model" refers to an artificial intelligence model that generates new text or information based on input data.

[1774] "Means for generating differentiating talk" refers to a method of using a generative AI model to generate a speaking style or sales pitch to propose a new company's plan that is differentiated from the plans and services of other companies.

[1775] "Means for outputting the generated differentiated talk" refers to a method for providing a user with the differentiated talk created by the generative AI model.

[1776] "Means of generating countertalk" refers to the way in which you generate speech or sales pitches to counter an announcement of a new plan by another company.

[1777] The present invention is a system in which a server, a terminal, and a user work together, and is implemented as follows.

[1778] System Configuration

[1779] The server has the following main functions:

[1780] 1. Data collection function:

[1781] The server collects plan and service information from other companies. This information is obtained through web scraping tools (e.g. BeautifulSoup, Scrapy) and APIs.

[1782] 2. Data pre-processing function:

[1783] The server preprocesses the collected data, specifically cleaning the text data (removing unnecessary characters and tags) and normalizing it (unifying uppercase and lowercase letters, converting special characters).

[1784] 3. Input functions for generative AI model:

[1785] The server inputs the preprocessed data into a generative AI model (e.g., GPT-4 (registered trademark)) by providing instructions to the model using prompt sentences.

[1786] 4. Differentiated Talk Generation Function:

[1787] The generative AI model generates talk to differentiate a company's new plan based on the input information about other companies' plans.

[1788] 5. Result output function:

[1789] The server outputs the generated differentiated token and transmits it to the terminal.

[1790] Hardware and Software Used

[1791] Web scraping tools: BeautifulSoup, Scrapy

[1792] Generative AI model: GPT-4 (registered trademark)

[1793] Data preprocessing tools: Python standard library and NLP libraries (e.g. NLTK, spaCy)

[1794] Examples

[1795] Example 1:

[1796] Other companies' plan information: "Unlimited calls for 1,000 yen per month"

[1797] Generated differentiated talk: "Unlimited calls and unlimited data for just 500 yen per month."

[1798] Example prompt:

[1799] "The other company's plan is 1,000 yen per month with unlimited calls. Please generate a chat to differentiate our new plan from the others."

[1800] How it works

[1801] 1. Data collection:

[1802] The server accesses the specified URL and retrieves the HTML content. It then uses BeautifulSoup to parse the HTML and extract the necessary information (plan name, price, service details, etc.).

[1803] 2. Data preprocessing:

[1804] The server removes HTML tags from the retrieved text data, removes extra spaces and line breaks, converts all characters to lower case, and converts special characters to their standard form.

[1805] 3. Input to the generative AI model:

[1806] The server inputs the preprocessed data into a generative AI model, providing instructions to the model using prompts.

[1807] 4. Generating Differentiated Talk:

[1808] The generative AI model analyzes information about other companies' plans and generates talk to propose new, competitive plans.

[1809] 5. Result output:

[1810] The server saves the chat received from the generative AI model in text format and sends it to the user's device. The user can review the chat and use it to propose new plans for their company.

[1811] In this way, the server utilizes generative AI models to provide a system that efficiently generates differentiated talks from other companies.

[1812] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1813] Step 1:

[1814] The server collects plan and service information from other companies.

[1815] Input: Third-party website URL or API endpoint.

[1816] Specific operation: The server accesses the specified URL and retrieves the HTML content. Then, it parses the HTML using BeautifulSoup or Scrapy and extracts the necessary information (plan name, price, service content, etc.).

[1817] Output: A list of extracted plans and services from other companies.

[1818] Step 2:

[1819] The server pre-processes the collected data.

[1820] Input: Third-party plan / service information collected in step 1.

[1821] What happens: The server removes HTML tags from the retrieved text, removes extra spaces and line breaks, converts all characters to lower case, and converts special characters to their standard form.

[1822] Output: Preprocessed and clean text data.

[1823] Step 3:

[1824] The server inputs the preprocessed data into a generative AI model.

[1825] Input: The clean text data preprocessed in step 2.

[1826] Specific operation: The server inputs the preprocessed data into a generative AI model (e.g., GPT-4 (registered trademark)). At this time, it gives instructions to the model using a prompt sentence. For example, it uses a prompt sentence such as, "Competitor's plans are 1,000 yen per month with unlimited calls. In contrast, please generate a chat to differentiate our new plan."

[1827] Output: The data and prompts fed into the generative AI model.

[1828] Step 4:

[1829] The generative AI model generates talk to differentiate a company's new plan based on the input information about other companies' plans.

[1830] Input: The data and prompts fed into the generative AI model in step 3.

[1831] Specific operation: The generative AI model analyzes the plan information of other companies and generates a message to propose a new competitive plan. For example, it generates a message saying, "Unlimited calls and unlimited data for 500 yen per month."

[1832] Output: The generated differentiation talk.

[1833] Step 5:

[1834] The server outputs the generated differentiated token and transmits it to the terminal.

[1835] Input: Differentiated tokens generated in step 4.

[1836] Specific operation: The server saves the chat received from the generative AI model in text format and sends it to the user's device. The user can review the chat and use it to propose new plans for their company.

[1837] Output: Differentiated token sent to user's device.

[1838] (Application example 2)

[1839] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server", and the headset type terminal 314 will be referred to as a "terminal".

[1840] In conventional systems, it is difficult to generate proposal talks for new plans to differentiate from competitors, and there is a lack of means for operators of online shopping sites to effectively differentiate from competitors. In addition, there is a lack of a way to easily share the generated proposal talks, which makes it difficult to respond quickly.

[1841] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1842] In this invention, the server includes a means for making an AI learn all plans and services of the company and other companies, and talks of the company's excellent crew, a means for generating talks that differentiate the company from other companies when the company issues a new plan, a means for generating counter talk when another company issues a new plan, a means for receiving a competitor's plan as an input and generating a proposal talk for the differentiated new plan, and a means for displaying the generated proposal talk to users and making it easy to share. This makes it possible to quickly and effectively generate and share proposal talks for new plans to differentiate the company from competitors.

[1843] "Plan" refers to the content and pricing of services or products offered by a company.

[1844] "Service" refers to any intangible value or benefit provided to a customer.

[1845] "Talk" refers to explanations or proposals given to customers in sales and marketing.

[1846] "AI" is an abbreviation for artificial intelligence, and refers to the technology that allows machines to imitate human intelligence to learn and make inferences.

[1847] "Differentiation talk" refers to explanations and proposals that emphasize the advantages of your company compared to your competitors' plans and services.

[1848] "Countertalk" refers to explanations or proposals made to counter a new plan announced by a competitor.

[1849] "Competitors" refers to other companies in the same market that offer similar plans or services.

[1850] A "proposal talk" refers to an explanation or suggestion given when proposing a new plan or service to a customer.

[1851] "User" refers to individuals or companies that use this system.

[1852] "Sharing" refers to sharing generated information or data with other users.

[1853] A system for implementing the present invention includes a server, a user terminal, and a generative AI model. The server receives input from the user terminal and generates differentiated talk and counter talk using the generative AI model. Specific embodiments are described below.

[1854] System Configuration

[1855] 1. Server

[1856] The server includes a database, an AI model, and an API. The database stores the plans and services of the company and other companies, and the talks of outstanding crew members. The AI ​​model is used to learn from these data and generate differentiated talks and counter talks. The API provides an interface for communication with user terminals.

[1857] 2. User terminal

[1858] The user terminal is a device such as a smartphone or a PC that is used by the user to input competitors' plans and check the generated tokens. The user terminal communicates with the server to send and receive the necessary data.

[1859] 3. Generative AI Models

[1860] Generative AI models use advanced natural language processing models such as OpenAI's GPT-3, which generate differentiation and countertalk based on a prompt.

[1861] Processing flow

[1862] 1. Data Entry

[1863] The user inputs a competitor's plan using the user terminal. For example, the user inputs a plan that offers "unlimited calls for 1,000 yen per month."

[1864] 2. Prompt generation

[1865] The server generates a prompt based on the input plan. An example of a prompt is "If the plan of another company is '1000 yen per month with unlimited calls,' please generate a proposal for a new differentiated plan."

[1866] 3. Creating a token

[1867] The server sends a prompt to the generative AI model to generate differentiated conversations. For example, the conversations generated might be something like, "Unlimited calls and unlimited data for 500 yen per month."

[1868] 4. View and share your results

[1869] The generated token is sent to the user's terminal and displayed to the user, who can then easily share the token.

[1870] Hardware and Software ...

Claims

1. A means to collect plan and service information from our company and other companies, as well as the contents of conversations between our excellent employees, and store them in a database; A means of preprocessing the accumulated data and training it using a generative AI model; A means for analyzing a user inquiry and generating an optimal answer using the generative AI model; A means for providing the generated answer to a user's terminal; A system including:

2. A means to have AI learn about our company's and other companies' plans and services, as well as the talk of our company's outstanding crew members, A way to generate a sales pitch that differentiates your company from other companies when launching a new plan. A way to generate counter talk when other companies release new plans, A means to suggest the best plans and services for the products users are considering purchasing, A means to provide users with the best options based on information on plans and services from our company and other companies, and The AI ​​will learn the conversations of the company's top crew members and provide optimal advice to users through chats. A system including:

3. A way to train artificial intelligence to learn all plans and services of our company and other companies, as well as the talk of our company's outstanding employees, A means of collecting information on plans and services of other companies; A means for pre-processing the collected data; means for inputting pre-processed data into said artificial intelligence; A means for generating a talk that differentiates the company from other companies using the artificial intelligence; A means for outputting the generated differentiated talk; A way to generate counter talk when other companies release new plans, A system including:

Images