system

Abstract

Provide a system. [Solution] a means for receiving questions from users; A means for generating an answer to the question using a generative AI; means for verifying the accuracy of the generated answers; A means for escalating any concerns raised by the verification; A means to provide the answer to the user after final confirmation by the sales representative; A system including:

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 an instruction sentence related to a description of 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] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In today's business environment, responding to numerous inquiries generated in corporate sales activities takes up a significant portion of sales representatives' man-hours, resulting in insufficient time for generating new business. Furthermore, there is a risk that incorrect or insufficient responses could reduce customer satisfaction. In such cases, there is a need for a system that automates inquiry responses and provides prompt and accurate answers while preventing incorrect responses. Furthermore, a system is needed that allows for appropriate confirmation and escalation when an unclear response is generated. This invention aims to solve these problems. [Means for solving the problem]

[0005] The present invention first comprises a means for receiving a question from a user and a means for generating an answer to the question using a generative AI. By providing a means for verifying the accuracy of the generated answer and a means for escalating any uncertainties that arise from the verification, incorrect answers can be prevented. The system further includes a means for providing an answer to the user after a final confirmation by a sales representative. This provides a system that achieves both speed and quality of response by automatically responding to routine inquiries and manually confirming unclear content. Furthermore, the generative AI retrieves information from an internal database and automatically escalates to a sales representative, enabling efficient and reliable inquiry responses.

[0006] "User" refers to any entity or individual that queries the system.

[0007] A "question" refers to a query for information or knowledge that a user provides to the system.

[0008] "Generative AI" refers to a system that uses artificial intelligence technology to automatically generate answers to questions from users.

[0009] "Answer" refers to the information generated by a generative AI in response to a question or the information provided to a user.

[0010] "Accuracy" refers to the state in which the answers generated by generative AI are factual and error-free.

[0011] "Validation" refers to the process for confirming the accuracy of the answers generated.

[0012] "Escalation" refers to the process of handing over any concerns or matters that require clarification to a sales representative.

[0013] "Sales Representative" means the entity or individual responsible for final review and necessary correction of escalated content.

[0014] "Final confirmation" refers to the process in which a sales representative checks the content of the answers generated by generative AI and makes appropriate corrections.

[0015] "Providing" refers to the process of communicating the final confirmed answer to the user.

[0016] "Internal database" refers to a database system used to manage information within a company.

[0017] "Automatically" refers to a state in which the system operates autonomously without the need for human intervention. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 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. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 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. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 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. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0019] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0020] First, the terms used in the following description will be explained.

[0021] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single 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), and an APU (Accelerated Processing Unit).

[0022] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0023] 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.

[0024] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0025] 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. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0026] [First embodiment]

[0027] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0028] 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.

[0029] 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 WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] 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 storage 50. 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.

[0031] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. 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.

[0032] 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 (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0034] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. 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 process 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.

[0036] 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.

[0037] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. 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.

[0038] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. 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."

[0039] This invention is a system that provides fast and accurate answers to inquiries from users in sales activities for corporations. Specifically, it uses generative AI to automatically generate answers, verifies their accuracy, and escalates them to a sales representative as necessary, thereby achieving both speed and quality in responses to users.

[0040] System configuration

[0041] The system consists of the following main components:

[0042] 1. User Interface

[0043] The terminal (user device) provides an interface for the user to enter questions. The user interface can be provided in a variety of forms, such as a web or mobile application.

[0044] 2. Server

[0045] Manages the core processes of the system, including receiving questions, generating answers, verifying answers, escalating, finalizing, and providing answers.

[0046] 3. Generative AI

[0047] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and constructing appropriate answers based on that information.

[0048] 4. Internal Database

[0049] This is an internal company database that is accessed by generative AI and contains product information, pricing information, inventory information, etc.

[0050] 5. Escalation function

[0051] The server has a function of transmitting the inquiry and the generated response to a sales representative when there is any concern about the generated response or when confirmation of accuracy is required.

[0052] 6. Salesperson Interface

[0053] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[0054] Program processing

[0055] 1. User inquiry submission

[0056] The user enters a question: "What is the price of product A?"

[0057] The terminal (user's device) sends this question to the server.

[0058] 2. Processing on the server

[0059] The server receives the query sent from the terminal.

[0060] The server sends the received questions to the generative AI.

[0061] 3. Generative AI Processing

[0062] Generative AI generates appropriate answers to questions.

[0063] For example, the answer "The price of product A is 10,000 yen" is retrieved from an internal database and generated.

[0064] The generated answer is sent back to the server.

[0065] 4. Response verification on the server

[0066] The server verifies the accuracy of the generated answer.

[0067] If the verification results show no problems, the answer is provided directly to the user.

[0068] If you are unsure about the content or need clarification, raise an escalation flag.

[0069] 5. Escalation Process

[0070] If the escalation flag is set, the server sends the inquiry and generated response to the sales representative.

[0071] 6. Final confirmation by sales representative

[0072] The sales representative receives the escalated content and checks and corrects it.

[0073] The final corrected answer is sent to the server.

[0074] 7. Providing answers to users

[0075] After the server receives the final response from the sales representative, it provides the response to the user.

[0076] The terminal displays to the user, "The price of product A is 10,000 yen."

[0077] Specific example explanation

[0078] Example 1: When confirmation is not required

[0079] 1. A user asks, "What is the price of product A?"

[0080] 2. The device sends this question to the server.

[0081] 3. The server receives the question and sends it to the generative AI.

[0082] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[0083] 5. The server receives this response and determines that there is no problem with the verification.

[0084] 6. The server provides the user with the answer, "The price of product A is 10,000 yen."

[0085] 7. The terminal displays the answer to the user.

[0086] Example 2: When confirmation is required

[0087] 1. A user asks, "What is the stock status of product B?"

[0088] 2. The device sends this question to the server.

[0089] 3. The server receives the question and sends it to the generative AI.

[0090] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[0091] 5. The server receives this response and sets the escalation flag.

[0092] 6. The server sends the inquiry and generated response to the sales representative.

[0093] 7. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[0094] 8. The server receives the final response and provides it to the user.

[0095] 9. The terminal displays to the user, "Product B is in stock."

[0096] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI and sales representatives.

[0097] The processing flow will be explained below.

[0098] Step 1:

[0099] The user enters a question, for example, "What is the price of product A?"

[0100] Step 2:

[0101] The terminal sends the user's question to the server.

[0102] Step 3:

[0103] The server receives the user's query.

[0104] Step 4:

[0105] The server sends the question to the generative AI.

[0106] Step 5:

[0107] The generative AI receives the question and retrieves relevant information from an internal database.

[0108] Step 6:

[0109] Generative AI generates an appropriate answer to the question, for example, "The price of product A is 10,000 yen."

[0110] Step 7:

[0111] The answer generated by the generative AI is sent to the server.

[0112] Step 8:

[0113] The server receives the answer from the generative AI.

[0114] Step 9:

[0115] The server verifies the accuracy of the generated answer.

[0116] Step 10:

[0117] If the server determines that confirmation is not necessary based on the verification results, it sends the answer to the user as is. For example, it sends the answer "The price of product A is 10,000 yen."

[0118] Step 11:

[0119] The user's terminal receives and displays the response.

[0120] Step 12:

[0121] If the server is unsure about the generated answer, it will raise an escalation flag.

[0122] Step 13:

[0123] The server sends the response flagged for escalation to the sales representative.

[0124] Step 14:

[0125] Sales reps receive and review escalated questions and answers.

[0126] Step 15:

[0127] After the sales representative checks the answer, he / she corrects the answer as necessary and sends the final answer to the server.

[0128] Step 16:

[0129] The server sends a final, confirmed response to the user, for example, "Product B is in stock."

[0130] Step 17:

[0131] The user's device receives and displays the final response, for example, "Product B is in stock."

[0132] The above is a specific flow of processing steps in this system.

[0133] Example 1

[0134] Next, a description will be given of 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."

[0135] In conventional corporate sales activities, it is difficult to provide quick and accurate responses to user inquiries, and there is a risk of delays in appropriate responses, especially when the inquiry is complex or requires accuracy. Furthermore, when human confirmation is required, the work of the person in charge increases and efficiency decreases, which raises concerns about lower customer satisfaction and a decline in the efficiency of sales activities. These issues need to be resolved.

[0136] 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.

[0137] In this invention, the server includes a means for receiving inquiries from users, a means for generating answers using a generative AI model, and a means for verifying the accuracy of the generated answers. This enables the automatic generation and provision of quick and accurate answers. Furthermore, if there are concerns about the accuracy of the answer, an escalation flag is set and the inquiry and the generated answer are sent to a sales representative, who then provides an appropriate answer to the user after final confirmation. This makes it possible to achieve both speed and quality in answers, thereby improving customer satisfaction and streamlining sales activities.

[0138] "User" means any person or entity that queries the System.

[0139] "Inquiry" refers to the act of a user providing information or asking a question to the system.

[0140] A "generative AI model" refers to an artificial intelligence program that uses natural language processing technology to automatically generate answers to user inquiries.

[0141] "Answer" refers to the information or response generated by a generative AI model in response to a user's inquiry.

[0142] "Accuracy" refers to the degree to which the generated answer is based on facts and correct information.

[0143] "Verification" refers to the process of checking whether the generated answers are accurate.

[0144] "Escalation Flag" refers to an identifying signal within the system that is set when there is concern about the accuracy of the answer generated.

[0145] "Salesperson" refers to the person who makes the final confirmation when an escalation flag is raised.

[0146] "Final review" refers to the process in which a sales representative reviews the answers provided by the generative AI model and makes corrections if necessary.

[0147] "Internal database" refers to a database that stores information that a generative AI model accesses to generate answers.

[0148] "Escalation" refers to the process of seeking confirmation from a sales representative on a generated response.

[0149] "Automatic" refers to the way the system operates autonomously, without the need for human intervention.

[0150] "System" refers to a set of components with a series of functions that accept user inquiries, generate answers using generative AI models, verify accuracy, escalate as necessary, and ultimately provide answers to users.

[0151] This invention is a system that provides fast and accurate answers to inquiries from users in corporate sales activities. Specifically, the system automatically generates answers using a generative AI model, verifies their accuracy, and escalates the inquiries to sales representatives as necessary, thereby achieving both speed and quality in responses to users.

[0152] System configuration

[0153] The system consists of the following main components:

[0154] 1. User Interface

[0155] The user's device provides an interface for the user to input questions. The user interface can be provided in a variety of forms, such as a web browser or a mobile application.

[0156] 2. Server

[0157] Manages the core processes of the system, including receiving questions, generating answers, verifying answers, escalating, finalizing, and providing answers.

[0158] 3. Generative AI Models

[0159] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and building appropriate answers based on that information.

[0160] 4. Internal Database

[0161] An internal company database accessed by generative AI models, containing product information, pricing information, inventory information, etc.

[0162] 5. Escalation function

[0163] The server has a function of transmitting the inquiry and the generated response to a sales representative when there is any concern about the generated response or when confirmation of accuracy is required.

[0164] 6. Salesperson Interface

[0165] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[0166] Program processing flow

[0167] The process flow for realizing this system is explained below:

[0168] 1. User inquiry submission

[0169] The user types a question into the terminal interface, such as "What is the price of product A?", and the terminal sends this question to the server in JSON format or another suitable data format.

[0170] 2. Receiving and analyzing questions

[0171] The server receives the inquiry sent from the terminal, analyzes the input data, and extracts the question, for example, "What is the price of product A?"

[0172] 3. Answer generation using generative AI models

[0173] The server sends the extracted question to the generative AI model, which then accesses its internal database and generates the optimal answer to the question. For example, it generates an answer such as "The price of product A is 10,000 yen."

[0174] 4. Response verification and escalation

[0175] The server uses its internal logic to verify the accuracy of the generated answer, and if it determines that the answer is inaccurate, it sets an escalation flag. Based on the escalation flag, the server sends the question and the generated answer to the sales representative.

[0176] 5. Final confirmation by sales representative

[0177] The sales representative receives the escalated content on the sales representative interface, checks the accuracy of the response, corrects the response as necessary, and sends the final, accurate information to the server.

[0178] 6. Providing a final response

[0179] The server receives the final response from the sales representative and provides it to the user. The user's terminal displays this final response, such as "The price of product A is 12,000 yen."

[0180] Specific example explanation

[0181] Example 1: No confirmation required

[0182] 1. A user asks, "What is the price of product A?"

[0183] 2. The device sends this question to the server.

[0184] 3. The server receives the question and sends it to the generative AI model.

[0185] 4. The generative AI model generates the answer, "The price of Product A is 10,000 yen."

[0186] 5. The server receives this response and determines that there is no problem with the verification.

[0187] 6. The server provides the user with the answer, "The price of product A is 10,000 yen."

[0188] 7. The device displays the answer to the user.

[0189] Example 2: When confirmation is required

[0190] 1. A user asks, "What is the availability of product B?"

[0191] 2. The device sends this question to the server.

[0192] 3. The server receives the question and sends it to the generative AI model.

[0193] 4. The generative AI model generates the answer, "The stock of product B is unknown. You may need to check."

[0194] 5. The server receives this response and sets the escalation flag.

[0195] 6. The server sends the inquiry and generated response to the sales representative.

[0196] 7. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[0197] 8. The server receives the final response and provides it to the user.

[0198] 9. The terminal displays to the user, "Product B is in stock."

[0199] Prompt Sentence Examples

[0200] "Please tell me the price of product A."

[0201] "Please let me know the stock status of product B."

[0202] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI models and sales representatives.

[0203] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0204] Step 1:

[0205] The user submits a query. The user enters a question into the terminal interface. For example, "What is the price of product A?" The terminal sends this question to the server as JSON format data. The input is the user's question, and the output is the JSON data sent to the server.

[0206] Step 2:

[0207] The server receives the question. The server receives the JSON data sent from the device, analyzes the data, and extracts the question content. For example, the information may include "Price of Product A." The input is the JSON data from the device, and the output is the extracted question content.

[0208] Step 3:

[0209] The server forwards the question to the generative AI model. The server generates an API request to pass the extracted question content to the generative AI model. The request includes the question content and the user's identification information. The input is the extracted question content and identification information, and the output is an API request to the generative AI model.

[0210] Step 4:

[0211] The generative AI model generates an answer. The generative AI model accesses an internal database to search for appropriate information for the question. For example, it obtains information such as "The price of product A is 10,000 yen" and generates an answer. The generated answer is returned to the server as an API response. The input is the API request, and the output is the generated answer.

[0212] Step 5:

[0213] The server receives and verifies the answer. The server receives the answer returned by the generative AI model and verifies its contents. For example, it uses internal logic to check the accuracy of the price information. The input is the generated answer, and the output is the verification result.

[0214] Step 6:

[0215] The server sets an escalation flag. If the server determines that the answer is inaccurate based on the validation results, it sets an escalation flag. For example, it sets a flag when the answer is ambiguous or requires additional confirmation. The input is the validation result, and the output is the escalation flag.

[0216] Step 7:

[0217] The server performs the escalation process. If the escalation flag is set, the server sends the generated response to the sales representative. This is done using an interface or email system dedicated to the sales representative. The input is the escalation flag and the generated response, and the output is the data to be sent to the sales representative.

[0218] Step 8:

[0219] The sales representative makes the final confirmation. The sales representative receives the escalated response via the interface and checks the contents. For example, they may correct the response to "The actual price of product A was 12,000 yen." The input is the escalated response, and the output is the corrected response.

[0220] Step 9:

[0221] The server receives the final answer. The server receives the revised answer from the sales representative and verifies its contents. The input is the revised answer, and the output is the confirmed final answer.

[0222] Step 10:

[0223] The user receives the final answer. The server sends the final answer to the user's device. The device displays to the user, "The price of product A is 12,000 yen." The input is the final answer, and the output is the information displayed to the user.

[0224] (Application example 1)

[0225] Next, a description will be given of Application 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."

[0226] Conventional customer support systems often have delayed responses to inquiries, which reduces customer satisfaction. Furthermore, in situations where immediate response is required in a physical store, it is difficult to provide a fast and accurate response. The present invention aims to solve these problems and provide a system that streamlines customer support in physical stores.

[0227] 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.

[0228] In this invention, the server includes means for receiving questions from users, means for generating answers using generative AI, means for verifying the accuracy of the answers, means for escalation, means for providing the answers to users after final confirmation by a sales representative, and means for displaying the answers on a terminal installed in a physical store. This makes it possible to provide quick and accurate answers to inquiries in a physical store and improve customer satisfaction.

[0229] "Means for receiving questions from users" is a function that allows the system to import questions entered by customers in a physical store using a tablet device, smartphone, etc.

[0230] "Generative AI" is an artificial intelligence model that automatically generates appropriate answers to questions from users, and often uses natural language processing technology.

[0231] "Means for verifying the accuracy of answers" is a function for checking whether the answers generated by generative AI are correct.

[0232] "Means for escalation" is a function that includes a process for raising the content of a generated answer to a human such as a sales representative if there are concerns about its accuracy.

[0233] The "means of providing the answer to the user after final confirmation by the sales representative" is a function for presenting the final answer that has been confirmed and corrected by the sales representative to the user.

[0234] "Means for displaying answers on a terminal installed in a physical store" is a function for displaying generated answers and corrected answers on a tablet terminal or display in a physical store.

[0235] This invention aims to improve the efficiency of customer support in brick-and-mortar stores. This system receives questions from users, uses generative AI to quickly and accurately generate answers, and verifies their appropriateness to improve customer satisfaction. The system is primarily composed of the following hardware and software:

[0236] System configuration and technologies used

[0237] 1. User Interface

[0238] The system uses tablets or smartphones installed in physical stores, allowing users to input specific questions.

[0239] 2. Server

[0240] It manages the core processing of the system, including receiving questions, generating answers using generative AI, verifying answers, escalating, final confirmation, and providing answers. It uses a general cloud server or on-premise server.

[0241] 3. Generative AI

[0242] It uses generative AI, such as the OpenAI (registered trademark) API, to automatically generate appropriate answers to user questions. It retrieves information from an internal database and generates prompts.

[0243] 4. Internal Database

[0244] We use a database that stores information necessary for sales activities, such as product information, price information, and inventory information.

[0245] 5. Escalation function

[0246] The server verifies the appropriateness of the generated answers and escalates to a sales representative if an incorrect answer is generated or requires clarification.

[0247] 6. Salesperson Interface

[0248] Provides an interface for sales representatives to review escalated content and make corrections or additions.

[0249] 7. Answer display function

[0250] The generated answers and the final answers confirmed and revised by the sales representative are displayed on tablet devices or displays in the physical store.

[0251] Processing flow

[0252] 1. User inputs a question

[0253] A user uses a tablet device in a physical store to type a question such as "Is this item in stock?"

[0254] 2. Receiving questions and generating answers using AI

[0255] The server receives the question and generates a prompt using the OpenAI API.

[0256] Example prompt: "User asks: Is this item in stock? Generate an appropriate answer."

[0257] 3. Response verification and escalation

[0258] The server verifies the generated answer, for example, "The item is not in stock. Please ask a store clerk for confirmation."

[0259] Escalate to sales if inaccurate or requires clarification.

[0260] 4. Final confirmation by sales representative

[0261] The sales representative checks the inventory status and corrects it to "This product is in stock."

[0262] 5. View Answers

[0263] The server receives the final answer and displays it on a tablet device in the physical store.

[0264] In this way, the present invention enables fast and accurate customer service in physical stores through collaboration between generative AI and sales representatives. Furthermore, the processing flow can be flexibly configured according to specific scenarios, aiming to improve customer satisfaction.

[0265] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0266] Step 1:

[0267] A user uses a tablet device in a physical store to enter a question. For example, they enter a question like "Is this product in stock?" The entered question is sent to the server by an application on the tablet device. The input data is in text format and is sent as an HTTP request to the server's question reception endpoint.

[0268] Step 2:

[0269] The server receives a question from a user. The received question is in text format and is first stored in a database. The server then generates a prompt to send the question to the generative AI. This prompt serves as an instruction for the generative AI model to generate an appropriate answer. For example, the format is "User question: Is this product in stock? Please generate an appropriate answer."

[0270] Step 3:

[0271] The generative AI receives the prompt sent from the server and generates an answer. The generative AI model (e.g., OpenAI GPT-3 (registered trademark)) generates an appropriate answer based on the prompt. This answer is generated using natural language processing technology and returned to the server in text format. An example of a generated answer is, "We cannot confirm the inventory. Please ask a store clerk to check."

[0272] Step 4:

[0273] The server verifies the accuracy of the answer returned by the generative AI. This verification is done by comparing it with information from an internal database. For example, if inventory information is not present in the database or if the AI's answer contains uncertainty, the server raises an escalation flag. If the answer is determined to be accurate, it proceeds to the next step.

[0274] Step 5:

[0275] If the server sets an escalation flag, it sends the question and the generated answer to the sales representative. The sending method is mainly via an internal API, and the sales representative's interface is notified. Here, the sales representative checks the question and the generated answer, and makes corrections or additions as necessary.

[0276] Step 6:

[0277] The sales representative receives the escalated information and checks the status. For example, they check the inventory information and enter a revised response into the system, such as "This product is in stock." The sales representative then sends the revised response to the server. This is mainly done via an internal API, and the server receives the revised response.

[0278] Step 7:

[0279] The server receives the final response from the sales representative and provides it to the user. The final response is sent to an application on a tablet device installed in the physical store and displayed to the user. The final response is then displayed on the user's screen, stating, "This product is in stock."

[0280] As described above, through the data processing and data calculation performed at each step, a system that provides quick and accurate answers to user inquiries is realized.

[0281] Furthermore, an emotion engine that estimates the user's emotion may be 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.

[0282] This invention is a system that provides prompt and accurate answers to inquiries from users in corporate sales activities. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the system aims to improve the user experience. Specifically, the emotion engine is used to detect the user's emotions and adjust the priority of responses and escalations based on those emotions.

[0283] System configuration

[0284] The system consists of the following main components:

[0285] 1. User Interface

[0286] The terminal (user device) provides an interface for the user to enter questions. The user interface can be provided in a variety of forms, such as a web or mobile application.

[0287] 2. Server

[0288] Manages the core processing of the system, including question receipt, answer generation, answer validation, escalation, final confirmation, answer delivery, and emotion recognition.

[0289] 3. Generative AI

[0290] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and constructing appropriate answers based on that information.

[0291] 4. Internal Database

[0292] This is an internal company database that is accessed by generative AI and contains product information, pricing information, inventory information, etc.

[0293] 5. Emotion Engine

[0294] This engine recognizes emotions from the user's input and writing style. For example, if the emotion engine detects negative emotions, it adjusts the responses generated by the generative AI and, if necessary, increases the priority of escalation.

[0295] 6. Escalation function

[0296] The server has a function of transmitting the inquiry and the generated answer to a sales representative when the user is unsure about the generated answer or when the user shows negative emotions.

[0297] 7. Salesperson Interface

[0298] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[0299] Program processing

[0300] 1. User inquiry submission

[0301] The user enters a question: "What is the price of product A?"

[0302] The terminal (user's device) sends this question to the server.

[0303] 2. Processing on the server

[0304] The server receives the user's query.

[0305] The server sends the question to the emotion engine to recognize the user's emotion.

[0306] 3. Generative AI Processing

[0307] The server sends the question to the generative AI.

[0308] Generative AI retrieves relevant information from an internal database and generates an appropriate answer to the question, for example, "The price of product A is 10,000 yen."

[0309] 4. Emotion Recognition by Emotion Engine

[0310] The emotion engine recognizes emotions from the user's input and writing style, for example, if the user expresses dissatisfaction or doubt.

[0311] 5. Response verification on the server

[0312] The server verifies the accuracy of the generated answer.

[0313] If the verification results show no problems, the answer is provided directly to the user.

[0314] 6. Response adjustment using emotion engine

[0315] The emotion engine adjusts the generative AI's responses based on the user's emotions, for example, using more polite language when negative emotions are recognized.

[0316] 7. Escalation Process

[0317] If the server is unsure about the generated answer or if the sentiment engine recognizes a negative sentiment, it raises an escalation flag.

[0318] The server sends the inquiry and generated response to the sales representative.

[0319] 8. Final confirmation by sales representative

[0320] Sales representatives receive escalated questions and answers and can review and correct them.

[0321] 9. Providing answers to users

[0322] After the server receives the final response from the sales representative, it provides the response to the user.

[0323] The terminal displays to the user, "The price of product A is 10,000 yen."

[0324] Specific example explanation

[0325] Example 1: When confirmation is not required

[0326] 1. A user asks, "What is the price of product A?"

[0327] 2. The device sends this question to the server.

[0328] 3. The server receives the question and sends it to the generative AI and emotion engine.

[0329] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[0330] 5. The emotion engine recognizes the user's emotions and does not detect negative emotions in particular.

[0331] 6. The server receives this response and determines that there is no problem with the verification.

[0332] 7. The server provides the user with the answer, "The price of product A is 10,000 yen."

[0333] 8. The device displays the answer to the user.

[0334] Example 2: When confirmation is required

[0335] 1. A user asks, "What is the stock status of product B?"

[0336] 2. The device sends this question to the server.

[0337] 3. The server receives the question and sends it to the generative AI and emotion engine.

[0338] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[0339] 5. The emotion engine recognizes when a user has anxiety.

[0340] 6. The server sets an escalation flag.

[0341] 7. The server sends the inquiry and generated response to the sales representative.

[0342] 8. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[0343] 9. The server provides the final answer to the user.

[0344] 10. The terminal displays to the user, "Product B is in stock."

[0345] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI, an emotion engine, and sales representatives, and also realizes flexible responses that correspond to the user's emotions.

[0346] The processing flow will be explained below.

[0347] Step 1:

[0348] The user enters a question: "What is the price of product A?"

[0349] Step 2:

[0350] The terminal sends a question to the server.

[0351] Step 3:

[0352] The server receives the user's query.

[0353] Step 4:

[0354] The server sends the question to the emotion engine.

[0355] Step 5:

[0356] The emotion engine analyzes the content of the question and recognizes the user's emotions, for example, assessing whether the user is anxious or suspicious based on the style and wording.

[0357] Step 6:

[0358] The server sends the question to the generative AI.

[0359] Step 7:

[0360] The generative AI analyzes the question and retrieves relevant information from an internal database.

[0361] Step 8:

[0362] Generative AI generates answers to questions, such as "The price of product A is 10,000 yen."

[0363] Step 9:

[0364] The answer generated by the generative AI is sent to the server.

[0365] Step 10:

[0366] The server receives the generated answer.

[0367] Step 11:

[0368] The server verifies the accuracy of the answer it receives, for example, by checking that it matches the latest information in its database.

[0369] Step 12:

[0370] The emotion engine adjusts responses based on the user's perceived emotions, for example adding the phrase "Don't worry" to the response if it detects that the user is anxious.

[0371] Step 13:

[0372] If the server determines based on the verification result that confirmation is not necessary, it sends the answer directly to the user.

[0373] Step 14:

[0374] The user's device receives and displays the answer. For example, it displays the answer "The price of product A is 10,000 yen."

[0375] Step 15:

[0376] If the server is unsure about the generated answer or if the sentiment engine detects a negative sentiment, it raises an escalation flag.

[0377] Step 16:

[0378] The server sends the response flagged for escalation to the sales representative.

[0379] Step 17:

[0380] Sales reps receive and review escalated questions and answers.

[0381] Step 18:

[0382] The sales representative checks the answer, corrects it if necessary, and sends the final answer to the server.

[0383] Step 19:

[0384] The server sends a final, confirmed response to the user, for example, "Product B is in stock."

[0385] Step 20:

[0386] The user's device receives and displays the final response, for example, "Product B is in stock."

[0387] The above is the flow of specific processing steps in a system that combines emotion engines.

[0388] Example 2

[0389] Next, a description will be given of 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."

[0390] In modern corporate sales activities, it is necessary to provide fast and accurate answers to user inquiries while also taking user emotions into consideration. However, existing systems lack accuracy and speed in their responses, and fail to properly reflect user emotions. This not only impairs the user experience, but also reduces sales efficiency.

[0391] 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.

[0392] In this invention, the server includes means for receiving questions from users, means for generating answers using generative AI, means for verifying the accuracy of the generated answers, emotion engine means for recognizing the user's emotions based on the content of the question, means for adjusting the content of the generated answers based on the emotions recognized by the emotion engine, means for escalating the question as necessary based on the results of the verification and emotion recognition, and means for providing the answer to the user after final confirmation by a sales representative. This enables the provision of fast and accurate answers and flexible responses based on the user's emotions.

[0393] "User" means a corporate or individual user who submits a question using the system.

[0394] A "means for receiving a query" is a component of the system for receiving and processing a query sent from a user's device.

[0395] "Generative AI" is an artificial intelligence engine that automatically generates answers to user questions based on an internal database.

[0396] "Means for generating answers" refers to the system's functionality for using generative AI to construct appropriate answers to users' questions.

[0397] "Means for verifying the accuracy of answers" refers to the system's process for checking that the answers generated contain appropriate and correct information.

[0398] The "emotion engine" is the engine of the system that recognizes emotions from the user's input content and writing style, and adjusts the response based on those emotions.

[0399] "Emotion recognition means" refers to the system's function of analyzing and evaluating the user's emotions through an emotion engine.

[0400] The "means for adjusting the content of the answer" is a function of the system that changes the wording and content of the generated answer based on the user's emotions recognized by the emotion engine.

[0401] "Means for escalation" refers to the system's function of passing on the inquiry and answer to a higher-level sales representative as needed, based on the generated answer and the user's sentiment.

[0402] "Sales Representative" means a company employee whose job is to provide final review of escalated inquiries and make corrections as necessary.

[0403] The "means of providing the answer to the user after final confirmation" is a system function that sends the final answer to the user after the sales representative has confirmed and corrected it.

[0404] This invention is a system for quickly and accurately responding to inquiries from users in corporate sales activities. This system receives questions from users, generates answers using generative AI, and uses an emotion engine to realize flexible responses that recognize and consider the user's emotions. Detailed embodiments of the present invention are described below.

[0405] System configuration

[0406] 1. User Interface

[0407] Users use a terminal (user device) to input questions using a web browser or mobile application. The user device provides an interface for sending the question data to the server.

[0408] 2. Server

[0409] The server manages the core processing of the system, specifically receiving questions, generating answers using generative AI, verifying the accuracy of the answers, recognizing emotions using an emotion engine, handling escalations, and providing answers to users.

[0410] 3. Generative AI

[0411] The generative AI works in conjunction with the server, retrieving relevant information from an internal database and automatically generating answers to user questions. The generative AI model uses natural language processing technology.

[0412] 4. Internal Database

[0413] Internal databases store necessary company information such as product information, price information, inventory information, etc. Generative AI accesses these databases and generates answers to questions.

[0414] 5. Emotion Engine

[0415] The emotion engine recognizes the emotion in a user's input and writing style, adjusts the content and wording of the response based on that emotion, and also detects when a user is expressing negative emotions and prioritizes escalation.

[0416] 6. Escalation function

[0417] If the server is concerned about the generated answer or if the emotion engine recognizes a negative emotion, it sets an escalation flag and sends the inquiry and the generated answer to the sales representative.

[0418] 7. Salesperson Interface

[0419] Sales representatives receive escalated content and have an interface for reviewing and correcting it, which consists of a web dashboard and a dedicated mobile application.

[0420] Specific examples

[0421] Example 1: No confirmation required

[0422] 1. A user asks, "What is the price of product A?"

[0423] 2. The device sends this question to the server.

[0424] 3. The server receives the question and sends it to the generative AI and emotion engine.

[0425] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[0426] 5. The emotion engine recognizes the user's emotions and does not detect negative emotions.

[0427] 6. The server receives this response and determines that there is no problem with the verification.

[0428] 7. The server provides the user with the answer, "The price of product A is 10,000 yen."

[0429] 8. The device displays the answer to the user.

[0430] Example 2: When confirmation is required

[0431] 1. A user asks, "What is the stock status of product B?"

[0432] 2. The device sends this question to the server.

[0433] 3. The server receives the question and sends it to the generative AI and emotion engine.

[0434] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[0435] 5. The emotion engine recognizes when a user is anxious.

[0436] 6. The server sets an escalation flag.

[0437] 7. The server sends the inquiry and generated response to the sales representative.

[0438] 8. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[0439] 9. The server provides the final answer to the user.

[0440] 10. The terminal displays to the user, "Product B is in stock."

[0441] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI, an emotion engine, and sales representatives, and also realizes flexible responses that correspond to the user's emotions.

[0442] Prompt Sentence Examples

[0443] "Please tell me the price of product A."

[0444] "Please let me know the stock status of product B."

[0445] "I would like to know the detailed specifications of product C."

[0446] By using the above, it is possible to respond appropriately and quickly to a wide variety of questions from users.

[0447] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0448] Step 1:

[0449] The user inputs a question such as "What is the price of product A?"

[0450] Specific actions: A user enters a question into an input form in a web browser or mobile application and presses the "Submit" button.

[0451] Input: Text input: "What is the price of product A?"

[0452] Output: The question is sent from the device to the server.

[0453] Step 2:

[0454] The terminal sends this question to the server.

[0455] Specific operation: The terminal sends the user's input to the server as an HTTP request.

[0456] Input: The question entered by the user.

[0457] Output: The question is sent to the server.

[0458] Step 3:

[0459] The server receives the user's query.

[0460] Specific operation: The server parses the HTTP request and extracts the question.

[0461] Input: The question sent from the device.

[0462] Output: The query is ready to be processed by the server.

[0463] Step 4:

[0464] The server sends the question content to the emotion engine and recognizes the emotion.

[0465] Specific operation: The server passes the question content to the emotion engine API and obtains the emotion score.

[0466] Input: The question.

[0467] Output: The emotion score returned by the emotion engine (e.g., happy, anger, sadness, etc.).

[0468] Step 5:

[0469] The server sends the question to the generative AI, which generates an answer.

[0470] How it works: The server calls the generative AI's API and inputs the question. The generative AI retrieves relevant information from its internal database and generates an appropriate answer.

[0471] Input: The question.

[0472] Output: The answer returned by the generative AI (e.g., "Product A costs 10,000 yen").

[0473] Step 6:

[0474] The server verifies the accuracy of the generated answer.

[0475] What happens: The server uses internal logic to check whether the generated answer is of the correct format and content, e.g., whether the price information is in the correct format.

[0476] Input: Answer from a generative AI.

[0477] Output: Answers verified for accuracy.

[0478] Step 7:

[0479] The server adjusts the response content based on the emotion recognized by the emotion engine.

[0480] Specific behavior: The server takes into account the emotion score from the emotion engine and adjusts the style and content of the response. For example, if a negative emotion is detected, it will change the wording to be more polite.

[0481] Input: Sentiment score, generated answer.

[0482] Output: The adjusted answer.

[0483] Step 8:

[0484] The server sets an escalation flag if necessary and sends the inquiry and generated response to a sales representative.

[0485] What it does: If the server is unsure about the generated answer or if the sentiment engine recognizes a negative sentiment, it sets an escalation flag and sends the inquiry and answer to a sales rep.

[0486] Input: A tailored response, and an escalation flag if necessary.

[0487] Output: Enquiry and response sent to sales representative.

[0488] Step 9:

[0489] Sales representatives receive escalated questions and answers and can review and correct them.

[0490] What happens next? Salespeople review details in a dedicated interface (web dashboard or mobile app) and enter revised responses as needed.

[0491] Input: Escalated question and answer.

[0492] Output: The final revised answer.

[0493] Step 10:

[0494] After receiving the final response from the sales representative, the server provides the response to the user.

[0495] Specific operation: The server sends the final answer to the terminal as an HTTP response.

[0496] Input: Final response from sales representative.

[0497] Output: The final answer sent to the user.

[0498] Step 11:

[0499] The terminal displays the final answer to the user.

[0500] Specific operation: The device receives the response from the server and displays the final answer in the user interface.

[0501] Input: The final response from the server.

[0502] Output: The final answer that is displayed to the user.

[0503] The above are the specific processing steps of this system, which allows for quick and accurate answers and flexible responses that reflect the user's emotions.

[0504] (Application example 2)

[0505] Next, a description will be given of Application 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."

[0506] Traditional corporate sales activities and customer support systems face challenges in providing prompt and accurate responses to user inquiries. Furthermore, a lack of response that reflects user emotions poses a risk of a poor user experience. In particular, if users with negative emotions are not properly addressed, customer satisfaction may decline, adversely affecting business growth. To solve these issues, a system that combines generative AI and emotion recognition technology and has an appropriate escalation function is needed.

[0507] 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.

[0508] In this invention, the server includes a means for receiving questions from users, a means for generating answers using a generative AI model, a means for using an emotion engine to recognize user emotions, a means for escalating if a negative emotion is recognized, and a means for providing an answer to the user after final confirmation by a sales representative. This enables quick and accurate responses to user inquiries and flexible responses according to emotions.

[0509] A "server" is a computer system that provides data and services in response to requests from other computers or devices.

[0510] "User" refers to any individual or legal entity that uses the system or services.

[0511] A "Question" is text or audio information provided by a User in the form of an inquiry or request for information.

[0512] A "generative AI model" is an artificial intelligence technology that automatically generates answers based on questions posed by users.

[0513] "Answer" refers to the information or solution provided by a generative AI model or a salesperson.

[0514] "Accuracy" refers to information or answers that are factual and free from errors.

[0515] An "emotion engine" is a technology that recognizes emotions from a user's questions or input, and provides appropriate responses and responses based on those emotions.

[0516] "Negative affect" refers to negative emotional states such as anxiety, doubt, anger, and sadness.

[0517] "Escalation" is the process of passing a difficult problem or situation to a higher level person or expert.

[0518] "Salesperson" is the staff member responsible for communicating with the user and confirming or correcting the final response.

[0519] An "internal database" is a data storage system installed within a company that a generative AI model accesses and retrieves information from.

[0520] This invention is a system for quickly and accurately responding to inquiries in corporate sales activities and customer support systems. In particular, it aims to improve the user experience by identifying the user's emotions and providing answers or escalations based on those emotions. The specific configuration and operation of this system are described in detail below.

[0521] System configuration

[0522] The system consists of the following main components:

[0523] 1. Server

[0524] The server receives questions from users, generates answers using a generative AI model, recognizes user emotions using an emotion engine, and escalates the situation as needed. The server also provides the final answer.

[0525] 2. Generative AI Models

[0526] The generative AI model retrieves information from an internal database and generates appropriate answers based on user questions. The generative AI model is built using natural language processing technology.

[0527] 3. Emotion Engine

[0528] The emotion engine recognizes emotions from user questions and inputs, especially negative emotions (e.g., anxiety or doubt), and adjusts the answers of the generative AI model.

[0529] 4. Escalation function

[0530] If the emotion engine recognizes a negative emotion or if the generated response is unsatisfactory, the server automatically escalates the inquiry and sends the generated response to a sales representative.

[0531] 5. Salesperson Interface

[0532] The sales representative receives the escalated questions and answers, checks and corrects them, and the final corrected answers are provided to the user.

[0533] Hardware and software used

[0534] Smartphone: Used to accept input from users and display results.

[0535] Server: Receives questions, generates answers, recognizes emotions, and handles escalation.

[0536] Generative AI model: Obtains information from an internal database and generates answers.

[0537] Emotion engine: Recognizes user emotions and tailors responses.

[0538] Specific examples

[0539] 1. Receiving user questions

[0540] A user uses a smartphone app to ask, "Can I return product X?" This question is sent to the server.

[0541] 2. Generating answers to questions

[0542] The generative AI model generates the answer, "Returns of product X are generally possible. Click here for details."

[0543] 3. User Emotion Recognition

[0544] The emotion engine recognizes the emotion of doubt from the user's question.

[0545] 4. Escalation

[0546] The server escalates based on the results of the emotion engine and sends the question and AI answer to the sales representative.

[0547] 5. Final confirmation and response

[0548] The sales representative will check the request and provide the final answer to the user via the server: "Returns are possible. If you have any questions, please contact our support center." The answer will be displayed on the smartphone.

[0549] Prompt Sentence Examples

[0550] Prompt for the AI ​​model to perform sentiment analysis:

[0551] Please analyze the following text and classify the emotion: 'Can I return product X?'

[0552] Prompt for generating an answer:

[0553] Generate a customer support response for the following question: 'Can I return product X?' using the company database.

[0554] In this way, the present invention integrates generative AI, an emotion engine, and an escalation function to realize an inquiry response system that can respond flexibly and appropriately according to the user's emotions.

[0555] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0556] Step 1:

[0557] The user inputs and sends a question using a device (smartphone app). An example of a user input is, "Can I return product X?" This input is sent to the server.

[0558] Step 2:

[0559] The server receives the question from the user, which is received in text format and stored internally for further processing.

[0560] Step 3:

[0561] The server sends the received question to the emotion engine, which recognizes the user's emotion from the question. For example, the emotion engine recognizes doubt from the question "Are returns possible?" The processing result is returned to the server.

[0562] Step 4:

[0563] The server sends the question and the emotion data obtained from the emotion engine to the generative AI model. The generative AI model retrieves the corresponding information from its internal database and generates an answer to the question. For example, it generates an answer such as, "Returns of product X are usually possible. Click here for details." This answer is then returned to the server.

[0564] Step 5:

[0565] The server verifies the answer obtained from the generative AI model. During this verification process, data from an internal database is referenced to confirm the accuracy and appropriateness of the answer. If there are no concerns as a result of the verification, the server proceeds to the next step.

[0566] Step 6:

[0567] The server uses data from the emotion engine to determine if the user has a negative emotion, and if a negative emotion is recognized, the server sets an escalation flag.

[0568] Step 7:

[0569] If a negative sentiment is detected, the server automatically escalates the query and sends the generated response to a sales representative.

[0570] Step 8:

[0571] The sales representative checks the received question and the AI-generated answer and modifies it as necessary. For example, if the answer is modified to "Returns are possible. If you have any questions, please contact our support center," the modified answer is returned to the server.

[0572] Step 9:

[0573] The server receives the final confirmed answer from the sales representative and provides it to the user. This answer is displayed on the device (smartphone app) and notified to the user.

[0574] 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 audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio 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 audio data.

[0575] 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 engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with 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.

[0576] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0577] [Second embodiment]

[0578] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0579] 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.

[0580] 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 WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0581] 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.

[0582] 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 audio in accordance with instructions from the processor 46.

[0583] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and 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 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0584] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0585] Fig. 4 shows an example of the 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.

[0586] 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.

[0587] 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.

[0588] 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.

[0589] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0590] This invention is a system that provides fast and accurate answers to inquiries from users in sales activities for corporations. Specifically, it uses generative AI to automatically generate answers, verifies their accuracy, and escalates them to a sales representative as necessary, thereby achieving both speed and quality in responses to users.

[0591] System configuration

[0592] The system consists of the following main components:

[0593] 1. User Interface

[0594] The terminal (user device) provides an interface for the user to enter questions. The user interface can be provided in a variety of forms, such as a web or mobile application.

[0595] 2. Server

[0596] Manages the core processes of the system, including receiving questions, generating answers, verifying answers, escalating, finalizing, and providing answers.

[0597] 3. Generative AI

[0598] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and constructing appropriate answers based on that information.

[0599] 4. Internal Database

[0600] This is an internal company database that is accessed by generative AI and contains product information, pricing information, inventory information, etc.

[0601] 5. Escalation function

[0602] The server has a function of transmitting the inquiry and the generated response to a sales representative when there is any concern about the generated response or when confirmation of accuracy is required.

[0603] 6. Salesperson Interface

[0604] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[0605] Program processing

[0606] 1. User inquiry submission

[0607] The user enters a question: "What is the price of product A?"

[0608] The terminal (user's device) sends this question to the server.

[0609] 2. Processing on the server

[0610] The server receives the query sent from the terminal.

[0611] The server sends the received questions to the generative AI.

[0612] 3. Generative AI Processing

[0613] Generative AI generates appropriate answers to questions.

[0614] For example, the answer "The price of product A is 10,000 yen" is retrieved from an internal database and generated.

[0615] The generated answer is sent back to the server.

[0616] 4. Response verification on the server

[0617] The server verifies the accuracy of the generated answer.

[0618] If the verification results show no problems, the answer is provided directly to the user.

[0619] If you are unsure about the content or need clarification, raise an escalation flag.

[0620] 5. Escalation Process

[0621] If the escalation flag is set, the server sends the inquiry and generated response to the sales representative.

[0622] 6. Final confirmation by sales representative

[0623] The sales representative receives the escalated content and checks and corrects it.

[0624] The final corrected answer is sent to the server.

[0625] 7. Providing answers to users

[0626] After the server receives the final response from the sales representative, it provides the response to the user.

[0627] The terminal displays to the user, "The price of product A is 10,000 yen."

[0628] Specific example explanation

[0629] Example 1: When confirmation is not required

[0630] 1. A user asks, "What is the price of product A?"

[0631] 2. The device sends this question to the server.

[0632] 3. The server receives the question and sends it to the generative AI.

[0633] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[0634] 5. The server receives this response and determines that there is no problem with the verification.

[0635] 6. The server provides the user with the answer, "The price of product A is 10,000 yen."

[0636] 7. The terminal displays the answer to the user.

[0637] Example 2: When confirmation is required

[0638] 1. A user asks, "What is the stock status of product B?"

[0639] 2. The device sends this question to the server.

[0640] 3. The server receives the question and sends it to the generative AI.

[0641] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[0642] 5. The server receives this response and sets the escalation flag.

[0643] 6. The server sends the inquiry and generated response to the sales representative.

[0644] 7. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[0645] 8. The server receives the final response and provides it to the user.

[0646] 9. The terminal displays to the user, "Product B is in stock."

[0647] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI and sales representatives.

[0648] The processing flow will be explained below.

[0649] Step 1:

[0650] The user enters a question, for example, "What is the price of product A?"

[0651] Step 2:

[0652] The terminal sends the user's question to the server.

[0653] Step 3:

[0654] The server receives the user's query.

[0655] Step 4:

[0656] The server sends the question to the generative AI.

[0657] Step 5:

[0658] The generative AI receives the question and retrieves relevant information from an internal database.

[0659] Step 6:

[0660] Generative AI generates an appropriate answer to the question, for example, "The price of product A is 10,000 yen."

[0661] Step 7:

[0662] The answer generated by the generative AI is sent to the server.

[0663] Step 8:

[0664] The server receives the answer from the generative AI.

[0665] Step 9:

[0666] The server verifies the accuracy of the generated answer.

[0667] Step 10:

[0668] If the server determines that confirmation is not necessary based on the verification results, it sends the answer to the user as is. For example, it sends the answer "The price of product A is 10,000 yen."

[0669] Step 11:

[0670] The user's terminal receives and displays the response.

[0671] Step 12:

[0672] If the server is unsure about the generated answer, it will raise an escalation flag.

[0673] Step 13:

[0674] The server sends the response flagged for escalation to the sales representative.

[0675] Step 14:

[0676] Sales reps receive and review escalated questions and answers.

[0677] Step 15:

[0678] After the sales representative checks the answer, he / she corrects the answer as necessary and sends the final answer to the server.

[0679] Step 16:

[0680] The server sends a final, confirmed response to the user, for example, "Product B is in stock."

[0681] Step 17:

[0682] The user's device receives and displays the final response, for example, "Product B is in stock."

[0683] The above is a specific flow of processing steps in this system.

[0684] Example 1

[0685] Next, a description will be given of 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."

[0686] In conventional corporate sales activities, it is difficult to provide quick and accurate responses to user inquiries, and there is a risk of delays in appropriate responses, especially when the inquiry is complex or requires accuracy. Furthermore, when human confirmation is required, the work of the person in charge increases and efficiency decreases, which raises concerns about lower customer satisfaction and a decline in the efficiency of sales activities. These issues need to be resolved.

[0687] 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.

[0688] In this invention, the server includes a means for receiving inquiries from users, a means for generating answers using a generative AI model, and a means for verifying the accuracy of the generated answers. This enables the automatic generation and provision of quick and accurate answers. Furthermore, if there are concerns about the accuracy of the answer, an escalation flag is set and the inquiry and the generated answer are sent to a sales representative, who then provides an appropriate answer to the user after final confirmation. This makes it possible to achieve both speed and quality in answers, thereby improving customer satisfaction and streamlining sales activities.

[0689] "User" means any person or entity that queries the System.

[0690] "Inquiry" refers to the act of a user providing information or asking a question to the system.

[0691] A "generative AI model" refers to an artificial intelligence program that uses natural language processing technology to automatically generate answers to user inquiries.

[0692] "Answer" refers to the information or response generated by a generative AI model in response to a user's inquiry.

[0693] "Accuracy" refers to the degree to which the generated answer is based on facts and correct information.

[0694] "Verification" refers to the process of checking whether the generated answers are accurate.

[0695] "Escalation Flag" refers to an identifying signal within the system that is set when there is concern about the accuracy of the answer generated.

[0696] "Salesperson" refers to the person who makes the final confirmation when an escalation flag is raised.

[0697] "Final review" refers to the process in which a sales representative reviews the answers provided by the generative AI model and makes corrections if necessary.

[0698] "Internal database" refers to a database that stores information that a generative AI model accesses to generate answers.

[0699] "Escalation" refers to the process of seeking confirmation from a sales representative on a generated response.

[0700] "Automatic" refers to the way the system operates autonomously, without the need for human intervention.

[0701] "System" refers to a set of components with a series of functions that accept user inquiries, generate answers using generative AI models, verify accuracy, escalate as necessary, and ultimately provide answers to users.

[0702] This invention is a system that provides fast and accurate answers to inquiries from users in corporate sales activities. Specifically, the system automatically generates answers using a generative AI model, verifies their accuracy, and escalates the inquiries to sales representatives as necessary, thereby achieving both speed and quality in responses to users.

[0703] System configuration

[0704] The system consists of the following main components:

[0705] 1. User Interface

[0706] The user's device provides an interface for the user to input questions. The user interface can be provided in a variety of forms, such as a web browser or a mobile application.

[0707] 2. Server

[0708] Manages the core processes of the system, including receiving questions, generating answers, verifying answers, escalating, finalizing, and providing answers.

[0709] 3. Generative AI Models

[0710] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and building appropriate answers based on that information.

[0711] 4. Internal Database

[0712] An internal company database accessed by generative AI models, containing product information, pricing information, inventory information, etc.

[0713] 5. Escalation function

[0714] The server has a function of transmitting the inquiry and the generated response to a sales representative when there is any concern about the generated response or when confirmation of accuracy is required.

[0715] 6. Salesperson Interface

[0716] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[0717] Program processing flow

[0718] The process flow for realizing this system is explained below:

[0719] 1. User inquiry submission

[0720] The user types a question into the terminal interface, such as "What is the price of product A?", and the terminal sends this question to the server in JSON format or another suitable data format.

[0721] 2. Receiving and analyzing questions

[0722] The server receives the inquiry sent from the terminal, analyzes the input data, and extracts the question, for example, "What is the price of product A?"

[0723] 3. Answer generation using generative AI models

[0724] The server sends the extracted question to the generative AI model, which then accesses its internal database and generates the optimal answer to the question. For example, it generates an answer such as "The price of product A is 10,000 yen."

[0725] 4. Response verification and escalation

[0726] The server uses its internal logic to verify the accuracy of the generated answer, and if it determines that the answer is inaccurate, it sets an escalation flag. Based on the escalation flag, the server sends the question and the generated answer to the sales representative.

[0727] 5. Final confirmation by sales representative

[0728] The sales representative receives the escalated content on the sales representative interface, checks the accuracy of the response, corrects the response as necessary, and sends the final, accurate information to the server.

[0729] 6. Providing a final response

[0730] The server receives the final response from the sales representative and provides it to the user. The user's terminal displays this final response, such as "The price of product A is 12,000 yen."

[0731] Specific example explanation

[0732] Example 1: No confirmation required

[0733] 1. A user asks, "What is the price of product A?"

[0734] 2. The device sends this question to the server.

[0735] 3. The server receives the question and sends it to the generative AI model.

[0736] 4. The generative AI model generates the answer, "The price of Product A is 10,000 yen."

[0737] 5. The server receives this response and determines that there is no problem with the verification.

[0738] 6. The server provides the user with the answer, "The price of product A is 10,000 yen."

[0739] 7. The device displays the answer to the user.

[0740] Example 2: When confirmation is required

[0741] 1. A user asks, "What is the availability of product B?"

[0742] 2. The device sends this question to the server.

[0743] 3. The server receives the question and sends it to the generative AI model.

[0744] 4. The generative AI model generates the answer, "The stock of product B is unknown. You may need to check."

[0745] 5. The server receives this response and sets the escalation flag.

[0746] 6. The server sends the inquiry and generated response to the sales representative.

[0747] 7. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[0748] 8. The server receives the final response and provides it to the user.

[0749] 9. The terminal displays to the user, "Product B is in stock."

[0750] Prompt Sentence Examples

[0751] "Please tell me the price of product A."

[0752] "Please let me know the stock status of product B."

[0753] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI models and sales representatives.

[0754] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0755] Step 1:

[0756] The user submits a query. The user enters a question into the terminal interface. For example, "What is the price of product A?" The terminal sends this question to the server as JSON format data. The input is the user's question, and the output is the JSON data sent to the server.

[0757] Step 2:

[0758] The server receives the question. The server receives the JSON data sent from the device, analyzes the data, and extracts the question content. For example, the information may include "Price of Product A." The input is the JSON data from the device, and the output is the extracted question content.

[0759] Step 3:

[0760] The server forwards the question to the generative AI model. The server generates an API request to pass the extracted question content to the generative AI model. The request includes the question content and the user's identification information. The input is the extracted question content and identification information, and the output is an API request to the generative AI model.

[0761] Step 4:

[0762] The generative AI model generates an answer. The generative AI model accesses an internal database to search for appropriate information for the question. For example, it obtains information such as "The price of product A is 10,000 yen" and generates an answer. The generated answer is returned to the server as an API response. The input is the API request, and the output is the generated answer.

[0763] Step 5:

[0764] The server receives and verifies the answer. The server receives the answer returned by the generative AI model and verifies its contents. For example, it uses internal logic to check the accuracy of the price information. The input is the generated answer, and the output is the verification result.

[0765] Step 6:

[0766] The server sets an escalation flag. If the server determines that the answer is inaccurate based on the validation results, it sets an escalation flag. For example, it sets a flag when the answer is ambiguous or requires additional confirmation. The input is the validation result, and the output is the escalation flag.

[0767] Step 7:

[0768] The server performs the escalation process. If the escalation flag is set, the server sends the generated response to the sales representative. This is done using an interface or email system dedicated to the sales representative. The input is the escalation flag and the generated response, and the output is the data to be sent to the sales representative.

[0769] Step 8:

[0770] The sales representative makes the final confirmation. The sales representative receives the escalated response via the interface and checks the contents. For example, they may correct the response to "The actual price of product A was 12,000 yen." The input is the escalated response, and the output is the corrected response.

[0771] Step 9:

[0772] The server receives the final answer. The server receives the revised answer from the sales representative and verifies its contents. The input is the revised answer, and the output is the confirmed final answer.

[0773] Step 10:

[0774] The user receives the final answer. The server sends the final answer to the user's device. The device displays to the user, "The price of product A is 12,000 yen." The input is the final answer, and the output is the information displayed to the user.

[0775] (Application example 1)

[0776] Next, a description will be given of Application 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."

[0777] Conventional customer support systems often have delayed responses to inquiries, which reduces customer satisfaction. Furthermore, in situations where immediate response is required in a physical store, it is difficult to provide a fast and accurate response. The present invention aims to solve these problems and provide a system that streamlines customer support in physical stores.

[0778] 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.

[0779] In this invention, the server includes means for receiving questions from users, means for generating answers using generative AI, means for verifying the accuracy of the answers, means for escalation, means for providing the answers to users after final confirmation by a sales representative, and means for displaying the answers on a terminal installed in a physical store. This makes it possible to provide quick and accurate answers to inquiries in a physical store and improve customer satisfaction.

[0780] "Means for receiving questions from users" is a function that allows the system to import questions entered by customers in a physical store using a tablet device, smartphone, etc.

[0781] "Generative AI" is an artificial intelligence model that automatically generates appropriate answers to questions from users, and often uses natural language processing technology.

[0782] "Means for verifying the accuracy of answers" is a function for checking whether the answers generated by generative AI are correct.

[0783] "Means for escalation" is a function that includes a process for raising the content of a generated answer to a human such as a sales representative if there are concerns about its accuracy.

[0784] The "means of providing the answer to the user after final confirmation by the sales representative" is a function for presenting the final answer that has been confirmed and corrected by the sales representative to the user.

[0785] "Means for displaying answers on a terminal installed in a physical store" is a function for displaying generated answers and corrected answers on a tablet terminal or display in a physical store.

[0786] This invention aims to improve the efficiency of customer support in brick-and-mortar stores. This system receives questions from users, uses generative AI to quickly and accurately generate answers, and verifies their appropriateness to improve customer satisfaction. The system is primarily composed of the following hardware and software:

[0787] System configuration and technologies used

[0788] 1. User Interface

[0789] The system uses tablets or smartphones installed in physical stores, allowing users to input specific questions.

[0790] 2. Server

[0791] It manages the core processing of the system, including receiving questions, generating answers using generative AI, verifying answers, escalating, final confirmation, and providing answers. It uses a general cloud server or on-premise server.

[0792] 3. Generative AI

[0793] It uses generative AI, such as OpenAI's API, to automatically generate appropriate answers to user questions, retrieving information from an internal database and generating prompts.

[0794] 4. Internal Database

[0795] We use a database that stores information necessary for sales activities, such as product information, price information, and inventory information.

[0796] 5. Escalation function

[0797] The server verifies the appropriateness of the generated answers and escalates to a sales representative if an incorrect answer is generated or requires clarification.

[0798] 6. Salesperson Interface

[0799] Provides an interface for sales representatives to review escalated content and make corrections or additions.

[0800] 7. Answer display function

[0801] The generated answers and the final answers confirmed and revised by the sales representative are displayed on tablet devices or displays in the physical store.

[0802] Processing flow

[0803] 1. User inputs a question

[0804] A user uses a tablet device in a physical store to type a question such as "Is this item in stock?"

[0805] 2. Receiving questions and generating answers using AI

[0806] The server receives the question and generates a prompt using the OpenAI API.

[0807] Example prompt: "User asks: Is this item in stock? Generate an appropriate answer."

[0808] 3. Response verification and escalation

[0809] The server verifies the generated answer, for example, "The item is not in stock. Please ask a store clerk for confirmation."

[0810] Escalate to sales if inaccurate or requires clarification.

[0811] 4. Final confirmation by sales representative

[0812] The sales representative checks the inventory status and corrects it to "This product is in stock."

[0813] 5. View Answers

[0814] The server receives the final answer and displays it on a tablet device in the physical store.

[0815] In this way, the present invention enables fast and accurate customer service in physical stores through collaboration between generative AI and sales representatives. Furthermore, the processing flow can be flexibly configured according to specific scenarios, aiming to improve customer satisfaction.

[0816] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0817] Step 1:

[0818] A user uses a tablet device in a physical store to enter a question. For example, they enter a question like "Is this product in stock?" The entered question is sent to the server by an application on the tablet device. The input data is in text format and is sent as an HTTP request to the server's question reception endpoint.

[0819] Step 2:

[0820] The server receives a question from a user. The received question is in text format and is first stored in a database. The server then generates a prompt to send the question to the generative AI. This prompt serves as an instruction for the generative AI model to generate an appropriate answer. For example, the format is "User question: Is this product in stock? Please generate an appropriate answer."

[0821] Step 3:

[0822] The generative AI receives the prompt sent from the server and generates an answer. The generative AI model (e.g., OpenAI GPT-3) generates an appropriate answer based on the prompt. This answer is generated using natural language processing technology and returned to the server in text format. An example of a generated answer is, "We cannot confirm the item is in stock. Please ask a store clerk to check."

[0823] Step 4:

[0824] The server verifies the accuracy of the answer returned by the generative AI. This verification is done by comparing it with information from an internal database. For example, if inventory information is not present in the database or if the AI's answer contains uncertainty, the server raises an escalation flag. If the answer is determined to be accurate, it proceeds to the next step.

[0825] Step 5:

[0826] If the server sets an escalation flag, it sends the question and the generated answer to the sales representative. The sending method is mainly via an internal API, and the sales representative's interface is notified. Here, the sales representative checks the question and the generated answer, and makes corrections or additions as necessary.

[0827] Step 6:

[0828] The sales representative receives the escalated information and checks the status. For example, they check the inventory information and enter a revised response into the system, such as "This product is in stock." The sales representative then sends the revised response to the server. This is mainly done via an internal API, and the server receives the revised response.

[0829] Step 7:

[0830] The server receives the final response from the sales representative and provides it to the user. The final response is sent to an application on a tablet device installed in the physical store and displayed to the user. The final response is then displayed on the user's screen, stating, "This product is in stock."

[0831] As described above, through the data processing and data calculation performed at each step, a system that provides quick and accurate answers to user inquiries is realized.

[0832] Furthermore, 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.

[0833] This invention is a system that provides prompt and accurate answers to inquiries from users in corporate sales activities. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the system aims to improve the user experience. Specifically, the emotion engine is used to detect the user's emotions and adjust the priority of responses and escalations based on those emotions.

[0834] System configuration

[0835] The system consists of the following main components:

[0836] 1. User Interface

[0837] The terminal (user device) provides an interface for the user to enter questions. The user interface can be provided in a variety of forms, such as a web or mobile application.

[0838] 2. Server

[0839] Manages the core processing of the system, including question receipt, answer generation, answer validation, escalation, final confirmation, answer delivery, and emotion recognition.

[0840] 3. Generative AI

[0841] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and constructing appropriate answers based on that information.

[0842] 4. Internal Database

[0843] This is an internal company database that is accessed by generative AI and contains product information, pricing information, inventory information, etc.

[0844] 5. Emotion Engine

[0845] This engine recognizes emotions from the user's input and writing style. For example, if the emotion engine detects negative emotions, it adjusts the responses generated by the generative AI and, if necessary, increases the priority of escalation.

[0846] 6. Escalation function

[0847] The server has a function of transmitting the inquiry and the generated answer to a sales representative when the user is unsure about the generated answer or when the user shows negative emotions.

[0848] 7. Salesperson Interface

[0849] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[0850] Program processing

[0851] 1. User inquiry submission

[0852] The user enters a question: "What is the price of product A?"

[0853] The terminal (user's device) sends this question to the server.

[0854] 2. Processing on the server

[0855] The server receives the user's query.

[0856] The server sends the question to the emotion engine to recognize the user's emotion.

[0857] 3. Generative AI Processing

[0858] The server sends the question to the generative AI.

[0859] Generative AI retrieves relevant information from an internal database and generates an appropriate answer to the question, for example, "The price of product A is 10,000 yen."

[0860] 4. Emotion Recognition by Emotion Engine

[0861] The emotion engine recognizes emotions from the user's input and writing style, for example, if the user expresses dissatisfaction or doubt.

[0862] 5. Response verification on the server

[0863] The server verifies the accuracy of the generated answer.

[0864] If the verification results show no problems, the answer is provided directly to the user.

[0865] 6. Response adjustment using emotion engine

[0866] The emotion engine adjusts the generative AI's responses based on the user's emotions, for example, using more polite language when negative emotions are recognized.

[0867] 7. Escalation Process

[0868] If the server is unsure about the generated answer or if the sentiment engine recognizes a negative sentiment, it raises an escalation flag.

[0869] The server sends the inquiry and generated response to the sales representative.

[0870] 8. Final confirmation by sales representative

[0871] Sales representatives receive escalated questions and answers and can review and correct them.

[0872] 9. Providing answers to users

[0873] After the server receives the final response from the sales representative, it provides the response to the user.

[0874] The terminal displays to the user, "The price of product A is 10,000 yen."

[0875] Specific example explanation

[0876] Example 1: When confirmation is not required

[0877] 1. A user asks, "What is the price of product A?"

[0878] 2. The device sends this question to the server.

[0879] 3. The server receives the question and sends it to the generative AI and emotion engine.

[0880] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[0881] 5. The emotion engine recognizes the user's emotions and does not detect negative emotions in particular.

[0882] 6. The server receives this response and determines that there is no problem with the verification.

[0883] 7. The server provides the user with the answer, "The price of product A is 10,000 yen."

[0884] 8. The device displays the answer to the user.

[0885] Example 2: When confirmation is required

[0886] 1. A user asks, "What is the stock status of product B?"

[0887] 2. The device sends this question to the server.

[0888] 3. The server receives the question and sends it to the generative AI and emotion engine.

[0889] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[0890] 5. The emotion engine recognizes when a user has anxiety.

[0891] 6. The server sets an escalation flag.

[0892] 7. The server sends the inquiry and generated response to the sales representative.

[0893] 8. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[0894] 9. The server provides the final answer to the user.

[0895] 10. The terminal displays to the user, "Product B is in stock."

[0896] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI, an emotion engine, and sales representatives, and also realizes flexible responses that correspond to the user's emotions.

[0897] The processing flow will be explained below.

[0898] Step 1:

[0899] The user enters a question: "What is the price of product A?"

[0900] Step 2:

[0901] The terminal sends a question to the server.

[0902] Step 3:

[0903] The server receives the user's query.

[0904] Step 4:

[0905] The server sends the question to the emotion engine.

[0906] Step 5:

[0907] The emotion engine analyzes the content of the question and recognizes the user's emotions, for example, assessing whether the user is anxious or suspicious based on the style and wording.

[0908] Step 6:

[0909] The server sends the question to the generative AI.

[0910] Step 7:

[0911] The generative AI analyzes the question and retrieves relevant information from an internal database.

[0912] Step 8:

[0913] Generative AI generates answers to questions, such as "The price of product A is 10,000 yen."

[0914] Step 9:

[0915] The answer generated by the generative AI is sent to the server.

[0916] Step 10:

[0917] The server receives the generated answer.

[0918] Step 11:

[0919] The server verifies the accuracy of the answer it receives, for example, by checking that it matches the latest information in its database.

[0920] Step 12:

[0921] The emotion engine adjusts responses based on the user's perceived emotions, for example adding the phrase "Don't worry" to the response if it detects that the user is anxious.

[0922] Step 13:

[0923] If the server determines based on the verification result that confirmation is not necessary, it sends the answer directly to the user.

[0924] Step 14:

[0925] The user's device receives and displays the answer. For example, it displays the answer "The price of product A is 10,000 yen."

[0926] Step 15:

[0927] If the server is unsure about the generated answer or if the sentiment engine detects a negative sentiment, it raises an escalation flag.

[0928] Step 16:

[0929] The server sends the response flagged for escalation to the sales representative.

[0930] Step 17:

[0931] Sales reps receive and review escalated questions and answers.

[0932] Step 18:

[0933] The sales representative checks the answer, corrects it if necessary, and sends the final answer to the server.

[0934] Step 19:

[0935] The server sends a final, confirmed response to the user, for example, "Product B is in stock."

[0936] Step 20:

[0937] The user's device receives and displays the final response, for example, "Product B is in stock."

[0938] The above is the flow of specific processing steps in a system that combines emotion engines.

[0939] Example 2

[0940] Next, a description will be given of 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."

[0941] In modern corporate sales activities, it is necessary to provide fast and accurate answers to user inquiries while also taking user emotions into consideration. However, existing systems lack accuracy and speed in their responses, and fail to properly reflect user emotions. This not only impairs the user experience, but also reduces sales efficiency.

[0942] 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.

[0943] In this invention, the server includes means for receiving questions from users, means for generating answers using generative AI, means for verifying the accuracy of the generated answers, emotion engine means for recognizing the user's emotions based on the content of the question, means for adjusting the content of the generated answers based on the emotions recognized by the emotion engine, means for escalating the question as necessary based on the results of the verification and emotion recognition, and means for providing the answer to the user after final confirmation by a sales representative. This enables the provision of fast and accurate answers and flexible responses based on the user's emotions.

[0944] "User" means a corporate or individual user who submits a question using the system.

[0945] A "means for receiving a query" is a component of the system for receiving and processing a query sent from a user's device.

[0946] "Generative AI" is an artificial intelligence engine that automatically generates answers to user questions based on an internal database.

[0947] "Means for generating answers" refers to the system's functionality for using generative AI to construct appropriate answers to users' questions.

[0948] "Means for verifying the accuracy of answers" refers to the system's process for checking that the answers generated contain appropriate and correct information.

[0949] The "emotion engine" is the engine of the system that recognizes emotions from the user's input content and writing style, and adjusts the response based on those emotions.

[0950] "Emotion recognition means" refers to the system's function of analyzing and evaluating the user's emotions through an emotion engine.

[0951] The "means for adjusting the content of the answer" is a function of the system that changes the wording and content of the generated answer based on the user's emotions recognized by the emotion engine.

[0952] "Means for escalation" refers to the system's function of passing on the inquiry and answer to a higher-level sales representative as needed, based on the generated answer and the user's sentiment.

[0953] "Sales Representative" means a company employee whose job is to provide final review of escalated inquiries and make corrections as necessary.

[0954] The "means of providing the answer to the user after final confirmation" is a system function that sends the final answer to the user after the sales representative has confirmed and corrected it.

[0955] This invention is a system for quickly and accurately responding to inquiries from users in corporate sales activities. This system receives questions from users, generates answers using generative AI, and uses an emotion engine to realize flexible responses that recognize and consider the user's emotions. Detailed embodiments of the present invention are described below.

[0956] System configuration

[0957] 1. User Interface

[0958] Users use a terminal (user device) to input questions using a web browser or mobile application. The user device provides an interface for sending the question data to the server.

[0959] 2. Server

[0960] The server manages the core processing of the system, specifically receiving questions, generating answers using generative AI, verifying the accuracy of the answers, recognizing emotions using an emotion engine, handling escalations, and providing answers to users.

[0961] 3. Generative AI

[0962] The generative AI works in conjunction with the server, retrieving relevant information from an internal database and automatically generating answers to user questions. The generative AI model uses natural language processing technology.

[0963] 4. Internal Database

[0964] Internal databases store necessary company information such as product information, price information, inventory information, etc. Generative AI accesses these databases and generates answers to questions.

[0965] 5. Emotion Engine

[0966] The emotion engine recognizes the emotion in a user's input and writing style, adjusts the content and wording of the response based on that emotion, and also detects when a user is expressing negative emotions and prioritizes escalation.

[0967] 6. Escalation function

[0968] If the server is concerned about the generated answer or if the emotion engine recognizes a negative emotion, it sets an escalation flag and sends the inquiry and the generated answer to the sales representative.

[0969] 7. Salesperson Interface

[0970] Sales representatives receive escalated content and have an interface for reviewing and correcting it, which consists of a web dashboard and a dedicated mobile application.

[0971] Specific examples

[0972] Example 1: No confirmation required

[0973] 1. A user asks, "What is the price of product A?"

[0974] 2. The device sends this question to the server.

[0975] 3. The server receives the question and sends it to the generative AI and emotion engine.

[0976] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[0977] 5. The emotion engine recognizes the user's emotions and does not detect negative emotions.

[0978] 6. The server receives this response and determines that there is no problem with the verification.

[0979] 7. The server provides the user with the answer, "The price of product A is 10,000 yen."

[0980] 8. The device displays the answer to the user.

[0981] Example 2: When confirmation is required

[0982] 1. A user asks, "What is the stock status of product B?"

[0983] 2. The device sends this question to the server.

[0984] 3. The server receives the question and sends it to the generative AI and emotion engine.

[0985] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[0986] 5. The emotion engine recognizes when a user is anxious.

[0987] 6. The server sets an escalation flag.

[0988] 7. The server sends the inquiry and generated response to the sales representative.

[0989] 8. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[0990] 9. The server provides the final answer to the user.

[0991] 10. The terminal displays to the user, "Product B is in stock."

[0992] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI, an emotion engine, and sales representatives, and also realizes flexible responses that correspond to the user's emotions.

[0993] Prompt Sentence Examples

[0994] "Please tell me the price of product A."

[0995] "Please let me know the stock status of product B."

[0996] "I would like to know the detailed specifications of product C."

[0997] By using the above, it is possible to respond appropriately and quickly to a wide variety of questions from users.

[0998] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0999] Step 1:

[1000] The user inputs a question such as "What is the price of product A?"

[1001] Specific actions: A user enters a question into an input form in a web browser or mobile application and presses the "Submit" button.

[1002] Input: Text input: "What is the price of product A?"

[1003] Output: The question is sent from the device to the server.

[1004] Step 2:

[1005] The terminal sends this question to the server.

[1006] Specific operation: The terminal sends the user's input to the server as an HTTP request.

[1007] Input: The question entered by the user.

[1008] Output: The question is sent to the server.

[1009] Step 3:

[1010] The server receives the user's query.

[1011] Specific operation: The server parses the HTTP request and extracts the question.

[1012] Input: The question sent from the device.

[1013] Output: The query is ready to be processed by the server.

[1014] Step 4:

[1015] The server sends the question content to the emotion engine and recognizes the emotion.

[1016] Specific operation: The server passes the question content to the emotion engine API and obtains the emotion score.

[1017] Input: The question.

[1018] Output: The emotion score returned by the emotion engine (e.g., happy, anger, sadness, etc.).

[1019] Step 5:

[1020] The server sends the question to the generative AI, which generates an answer.

[1021] How it works: The server calls the generative AI's API and inputs the question. The generative AI retrieves relevant information from its internal database and generates an appropriate answer.

[1022] Input: The question.

[1023] Output: The answer returned by the generative AI (e.g., "Product A costs 10,000 yen").

[1024] Step 6:

[1025] The server verifies the accuracy of the generated answer.

[1026] What happens: The server uses internal logic to check whether the generated answer is of the correct format and content, e.g., whether the price information is in the correct format.

[1027] Input: Answer from a generative AI.

[1028] Output: Answers verified for accuracy.

[1029] Step 7:

[1030] The server adjusts the response content based on the emotion recognized by the emotion engine.

[1031] Specific behavior: The server takes into account the emotion score from the emotion engine and adjusts the style and content of the response. For example, if a negative emotion is detected, it will change the wording to be more polite.

[1032] Input: Sentiment score, generated answer.

[1033] Output: The adjusted answer.

[1034] Step 8:

[1035] The server sets an escalation flag if necessary and sends the inquiry and generated response to a sales representative.

[1036] What it does: If the server is unsure about the generated answer or if the sentiment engine recognizes a negative sentiment, it sets an escalation flag and sends the inquiry and answer to a sales rep.

[1037] Input: A tailored response, and an escalation flag if necessary.

[1038] Output: Enquiry and response sent to sales representative.

[1039] Step 9:

[1040] Sales representatives receive escalated questions and answers and can review and correct them.

[1041] What happens next? Salespeople review details in a dedicated interface (web dashboard or mobile app) and enter revised responses as needed.

[1042] Input: Escalated question and answer.

[1043] Output: The final revised answer.

[1044] Step 10:

[1045] After receiving the final response from the sales representative, the server provides the response to the user.

[1046] Specific operation: The server sends the final answer to the terminal as an HTTP response.

[1047] Input: Final response from sales representative.

[1048] Output: The final answer sent to the user.

[1049] Step 11:

[1050] The terminal displays the final answer to the user.

[1051] Specific operation: The device receives the response from the server and displays the final answer in the user interface.

[1052] Input: The final response from the server.

[1053] Output: The final answer that is displayed to the user.

[1054] The above are the specific processing steps of this system, which allows for quick and accurate answers and flexible responses that reflect the user's emotions.

[1055] (Application example 2)

[1056] Next, a description will be given of Application 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."

[1057] Traditional corporate sales activities and customer support systems face challenges in providing prompt and accurate responses to user inquiries. Furthermore, a lack of response that reflects user emotions poses a risk of a poor user experience. In particular, if users with negative emotions are not properly addressed, customer satisfaction may decline, adversely affecting business growth. To solve these issues, a system that combines generative AI and emotion recognition technology and has an appropriate escalation function is needed.

[1058] 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.

[1059] In this invention, the server includes a means for receiving questions from users, a means for generating answers using a generative AI model, a means for using an emotion engine to recognize user emotions, a means for escalating if a negative emotion is recognized, and a means for providing an answer to the user after final confirmation by a sales representative. This enables quick and accurate responses to user inquiries and flexible responses according to emotions.

[1060] A "server" is a computer system that provides data and services in response to requests from other computers or devices.

[1061] "User" refers to any individual or legal entity that uses the system or services.

[1062] A "Question" is text or audio information provided by a User in the form of an inquiry or request for information.

[1063] A "generative AI model" is an artificial intelligence technology that automatically generates answers based on questions posed by users.

[1064] "Answer" refers to the information or solution provided by a generative AI model or a salesperson.

[1065] "Accuracy" refers to information or answers that are factual and free from errors.

[1066] An "emotion engine" is a technology that recognizes emotions from a user's questions or input, and provides appropriate responses and responses based on those emotions.

[1067] "Negative affect" refers to negative emotional states such as anxiety, doubt, anger, and sadness.

[1068] "Escalation" is the process of passing a difficult problem or situation to a higher level person or expert.

[1069] "Salesperson" is the staff member responsible for communicating with the user and confirming or correcting the final response.

[1070] An "internal database" is a data storage system installed within a company that a generative AI model accesses and retrieves information from.

[1071] This invention is a system for quickly and accurately responding to inquiries in corporate sales activities and customer support systems. In particular, it aims to improve the user experience by identifying the user's emotions and providing answers or escalations based on those emotions. The specific configuration and operation of this system are described in detail below.

[1072] System configuration

[1073] The system consists of the following main components:

[1074] 1. Server

[1075] The server receives questions from users, generates answers using a generative AI model, recognizes user emotions using an emotion engine, and escalates the situation as needed. The server also provides the final answer.

[1076] 2. Generative AI Models

[1077] The generative AI model retrieves information from an internal database and generates appropriate answers based on user questions. The generative AI model is built using natural language processing technology.

[1078] 3. Emotion Engine

[1079] The emotion engine recognizes emotions from user questions and inputs, especially negative emotions (e.g., anxiety or doubt), and adjusts the answers of the generative AI model.

[1080] 4. Escalation function

[1081] If the emotion engine recognizes a negative emotion or if the generated response is unsatisfactory, the server automatically escalates the inquiry and sends the generated response to a sales representative.

[1082] 5. Salesperson Interface

[1083] The sales representative receives the escalated questions and answers, checks and corrects them, and the final corrected answers are provided to the user.

[1084] Hardware and software used

[1085] Smartphone: Used to accept input from users and display results.

[1086] Server: Receives questions, generates answers, recognizes emotions, and handles escalation.

[1087] Generative AI model: Obtains information from an internal database and generates answers.

[1088] Emotion engine: Recognizes user emotions and tailors responses.

[1089] Specific examples

[1090] 1. Receiving user questions

[1091] A user uses a smartphone app to ask, "Can I return product X?" This question is sent to the server.

[1092] 2. Generating answers to questions

[1093] The generative AI model generates the answer, "Returns of product X are generally possible. Click here for details."

[1094] 3. User Emotion Recognition

[1095] The emotion engine recognizes the emotion of doubt from the user's question.

[1096] 4. Escalation

[1097] The server escalates based on the results of the emotion engine and sends the question and AI answer to the sales representative.

[1098] 5. Final confirmation and response

[1099] The sales representative will check the request and provide the final answer to the user via the server: "Returns are possible. If you have any questions, please contact our support center." The answer will be displayed on the smartphone.

[1100] Prompt Sentence Examples

[1101] Prompt for the AI ​​model to perform sentiment analysis:

[1102] Please analyze the following text and classify the emotion: 'Can I return product X?'

[1103] Prompt for generating an answer:

[1104] Generate a customer support response for the following question: 'Can I return product X?' using the company database.

[1105] In this way, the present invention integrates generative AI, an emotion engine, and an escalation function to realize an inquiry response system that can respond flexibly and appropriately according to the user's emotions.

[1106] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1107] Step 1:

[1108] The user inputs and sends a question using a device (smartphone app). An example of a user input is, "Can I return product X?" This input is sent to the server.

[1109] Step 2:

[1110] The server receives the question from the user, which is received in text format and stored internally for further processing.

[1111] Step 3:

[1112] The server sends the received question to the emotion engine, which recognizes the user's emotion from the question. For example, the emotion engine recognizes doubt from the question "Are returns possible?" The processing result is returned to the server.

[1113] Step 4:

[1114] The server sends the question and the emotion data obtained from the emotion engine to the generative AI model. The generative AI model retrieves the corresponding information from its internal database and generates an answer to the question. For example, it generates an answer such as, "Returns of product X are usually possible. Click here for details." This answer is then returned to the server.

[1115] Step 5:

[1116] The server verifies the answer obtained from the generative AI model. During this verification process, data from an internal database is referenced to confirm the accuracy and appropriateness of the answer. If there are no concerns as a result of the verification, the server proceeds to the next step.

[1117] Step 6:

[1118] The server uses data from the emotion engine to determine if the user has a negative emotion, and if a negative emotion is recognized, the server sets an escalation flag.

[1119] Step 7:

[1120] If a negative sentiment is detected, the server automatically escalates the query and sends the generated response to a sales representative.

[1121] Step 8:

[1122] The sales representative checks the received question and the AI-generated answer and modifies it as necessary. For example, if the answer is modified to "Returns are possible. If you have any questions, please contact our support center," the modified answer is returned to the server.

[1123] Step 9:

[1124] The server receives the final confirmed answer from the sales representative and provides it to the user. This answer is displayed on the device (smartphone app) and notified to the user.

[1125] 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 audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio 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 audio data.

[1126] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with 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.

[1127] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1128] [Third embodiment]

[1129] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1130] 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.

[1131] 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 WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1132] 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.

[1133] 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 audio in accordance with instructions from the processor 46.

[1134] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and 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 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1135] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1136] Fig. 6 shows an example of the 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.

[1137] 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.

[1138] 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.

[1139] In the headset type terminal 314, a 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.

[1140] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[1141] This invention is a system that provides fast and accurate answers to inquiries from users in sales activities for corporations. Specifically, it uses generative AI to automatically generate answers, verifies their accuracy, and escalates them to a sales representative as necessary, thereby achieving both speed and quality in responses to users.

[1142] System configuration

[1143] The system consists of the following main components:

[1144] 1. User Interface

[1145] The terminal (user device) provides an interface for the user to enter questions. The user interface can be provided in a variety of forms, such as a web or mobile application.

[1146] 2. Server

[1147] Manages the core processes of the system, including receiving questions, generating answers, verifying answers, escalating, finalizing, and providing answers.

[1148] 3. Generative AI

[1149] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and constructing appropriate answers based on that information.

[1150] 4. Internal Database

[1151] This is an internal company database that is accessed by generative AI and contains product information, pricing information, inventory information, etc.

[1152] 5. Escalation function

[1153] The server has a function of transmitting the inquiry and the generated response to a sales representative when there is any concern about the generated response or when confirmation of accuracy is required.

[1154] 6. Salesperson Interface

[1155] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[1156] Program processing

[1157] 1. User inquiry submission

[1158] The user enters a question: "What is the price of product A?"

[1159] The terminal (user's device) sends this question to the server.

[1160] 2. Processing on the server

[1161] The server receives the query sent from the terminal.

[1162] The server sends the received questions to the generative AI.

[1163] 3. Generative AI Processing

[1164] Generative AI generates appropriate answers to questions.

[1165] For example, the answer "The price of product A is 10,000 yen" is retrieved from an internal database and generated.

[1166] The generated answer is sent back to the server.

[1167] 4. Response verification on the server

[1168] The server verifies the accuracy of the generated answer.

[1169] If the verification results show no problems, the answer is provided directly to the user.

[1170] If you are unsure about the content or need clarification, raise an escalation flag.

[1171] 5. Escalation Process

[1172] If the escalation flag is set, the server sends the inquiry and generated response to the sales representative.

[1173] 6. Final confirmation by sales representative

[1174] The sales representative receives the escalated content and checks and corrects it.

[1175] The final corrected answer is sent to the server.

[1176] 7. Providing answers to users

[1177] After the server receives the final response from the sales representative, it provides the response to the user.

[1178] The terminal displays to the user, "The price of product A is 10,000 yen."

[1179] Specific example explanation

[1180] Example 1: When confirmation is not required

[1181] 1. A user asks, "What is the price of product A?"

[1182] 2. The device sends this question to the server.

[1183] 3. The server receives the question and sends it to the generative AI.

[1184] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[1185] 5. The server receives this response and determines that there is no problem with the verification.

[1186] 6. The server provides the user with the answer, "The price of product A is 10,000 yen."

[1187] 7. The terminal displays the answer to the user.

[1188] Example 2: When confirmation is required

[1189] 1. A user asks, "What is the stock status of product B?"

[1190] 2. The device sends this question to the server.

[1191] 3. The server receives the question and sends it to the generative AI.

[1192] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[1193] 5. The server receives this response and sets the escalation flag.

[1194] 6. The server sends the inquiry and generated response to the sales representative.

[1195] 7. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[1196] 8. The server receives the final response and provides it to the user.

[1197] 9. The terminal displays to the user, "Product B is in stock."

[1198] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI and sales representatives.

[1199] The processing flow will be explained below.

[1200] Step 1:

[1201] The user enters a question, for example, "What is the price of product A?"

[1202] Step 2:

[1203] The terminal sends the user's question to the server.

[1204] Step 3:

[1205] The server receives the user's query.

[1206] Step 4:

[1207] The server sends the question to the generative AI.

[1208] Step 5:

[1209] The generative AI receives the question and retrieves relevant information from an internal database.

[1210] Step 6:

[1211] Generative AI generates an appropriate answer to the question, for example, "The price of product A is 10,000 yen."

[1212] Step 7:

[1213] The answer generated by the generative AI is sent to the server.

[1214] Step 8:

[1215] The server receives the answer from the generative AI.

[1216] Step 9:

[1217] The server verifies the accuracy of the generated answer.

[1218] Step 10:

[1219] If the server determines that confirmation is not necessary based on the verification results, it sends the answer to the user as is. For example, it sends the answer "The price of product A is 10,000 yen."

[1220] Step 11:

[1221] The user's terminal receives and displays the response.

[1222] Step 12:

[1223] If the server is unsure about the generated answer, it will raise an escalation flag.

[1224] Step 13:

[1225] The server sends the response flagged for escalation to the sales representative.

[1226] Step 14:

[1227] Sales reps receive and review escalated questions and answers.

[1228] Step 15:

[1229] After the sales representative checks the answer, he / she corrects the answer as necessary and sends the final answer to the server.

[1230] Step 16:

[1231] The server sends a final, confirmed response to the user, for example, "Product B is in stock."

[1232] Step 17:

[1233] The user's device receives and displays the final response, for example, "Product B is in stock."

[1234] The above is a specific flow of processing steps in this system.

[1235] Example 1

[1236] Next, a description will be given of 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."

[1237] In conventional corporate sales activities, it is difficult to provide quick and accurate responses to user inquiries, and there is a risk of delays in appropriate responses, especially when the inquiry is complex or requires accuracy. Furthermore, when human confirmation is required, the work of the person in charge increases and efficiency decreases, which raises concerns about lower customer satisfaction and a decline in the efficiency of sales activities. These issues need to be resolved.

[1238] 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.

[1239] In this invention, the server includes a means for receiving inquiries from users, a means for generating answers using a generative AI model, and a means for verifying the accuracy of the generated answers. This enables the automatic generation and provision of quick and accurate answers. Furthermore, if there are concerns about the accuracy of the answer, an escalation flag is set and the inquiry and the generated answer are sent to a sales representative, who then provides an appropriate answer to the user after final confirmation. This makes it possible to achieve both speed and quality in answers, thereby improving customer satisfaction and streamlining sales activities.

[1240] "User" means any person or entity that queries the System.

[1241] "Inquiry" refers to the act of a user providing information or asking a question to the system.

[1242] A "generative AI model" refers to an artificial intelligence program that uses natural language processing technology to automatically generate answers to user inquiries.

[1243] "Answer" refers to the information or response generated by a generative AI model in response to a user's inquiry.

[1244] "Accuracy" refers to the degree to which the generated answer is based on facts and correct information.

[1245] "Verification" refers to the process of checking whether the generated answers are accurate.

[1246] "Escalation Flag" refers to an identifying signal within the system that is set when there is concern about the accuracy of the answer generated.

[1247] "Salesperson" refers to the person who makes the final confirmation when an escalation flag is raised.

[1248] "Final review" refers to the process in which a sales representative reviews the answers provided by the generative AI model and makes corrections if necessary.

[1249] "Internal database" refers to a database that stores information that a generative AI model accesses to generate answers.

[1250] "Escalation" refers to the process of seeking confirmation from a sales representative on a generated response.

[1251] "Automatic" refers to the way the system operates autonomously, without the need for human intervention.

[1252] "System" refers to a set of components with a series of functions that accept user inquiries, generate answers using generative AI models, verify accuracy, escalate as necessary, and ultimately provide answers to users.

[1253] This invention is a system that provides fast and accurate answers to inquiries from users in corporate sales activities. Specifically, the system automatically generates answers using a generative AI model, verifies their accuracy, and escalates the inquiries to sales representatives as necessary, thereby achieving both speed and quality in responses to users.

[1254] System configuration

[1255] The system consists of the following main components:

[1256] 1. User Interface

[1257] The user's device provides an interface for the user to input questions. The user interface can be provided in a variety of forms, such as a web browser or a mobile application.

[1258] 2. Server

[1259] Manages the core processes of the system, including receiving questions, generating answers, verifying answers, escalating, finalizing, and providing answers.

[1260] 3. Generative AI Models

[1261] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and building appropriate answers based on that information.

[1262] 4. Internal Database

[1263] An internal company database accessed by generative AI models, containing product information, pricing information, inventory information, etc.

[1264] 5. Escalation function

[1265] The server has a function of transmitting the inquiry and the generated response to a sales representative when there is any concern about the generated response or when confirmation of accuracy is required.

[1266] 6. Salesperson Interface

[1267] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[1268] Program processing flow

[1269] The process flow for realizing this system is explained below:

[1270] 1. User inquiry submission

[1271] The user types a question into the terminal interface, such as "What is the price of product A?", and the terminal sends this question to the server in JSON format or another suitable data format.

[1272] 2. Receiving and analyzing questions

[1273] The server receives the inquiry sent from the terminal, analyzes the input data, and extracts the question, for example, "What is the price of product A?"

[1274] 3. Answer generation using generative AI models

[1275] The server sends the extracted question to the generative AI model, which then accesses its internal database and generates the optimal answer to the question. For example, it generates an answer such as "The price of product A is 10,000 yen."

[1276] 4. Response verification and escalation

[1277] The server uses its internal logic to verify the accuracy of the generated answer, and if it determines that the answer is inaccurate, it sets an escalation flag. Based on the escalation flag, the server sends the question and the generated answer to the sales representative.

[1278] 5. Final confirmation by sales representative

[1279] The sales representative receives the escalated content on the sales representative interface, checks the accuracy of the response, corrects the response as necessary, and sends the final, accurate information to the server.

[1280] 6. Providing a final response

[1281] The server receives the final response from the sales representative and provides it to the user. The user's terminal displays this final response, such as "The price of product A is 12,000 yen."

[1282] Specific example explanation

[1283] Example 1: No confirmation required

[1284] 1. A user asks, "What is the price of product A?"

[1285] 2. The device sends this question to the server.

[1286] 3. The server receives the question and sends it to the generative AI model.

[1287] 4. The generative AI model generates the answer, "The price of Product A is 10,000 yen."

[1288] 5. The server receives this response and determines that there is no problem with the verification.

[1289] 6. The server provides the user with the answer, "The price of product A is 10,000 yen."

[1290] 7. The device displays the answer to the user.

[1291] Example 2: When confirmation is required

[1292] 1. A user asks, "What is the availability of product B?"

[1293] 2. The device sends this question to the server.

[1294] 3. The server receives the question and sends it to the generative AI model.

[1295] 4. The generative AI model generates the answer, "The stock of product B is unknown. You may need to check."

[1296] 5. The server receives this response and sets the escalation flag.

[1297] 6. The server sends the inquiry and generated response to the sales representative.

[1298] 7. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[1299] 8. The server receives the final response and provides it to the user.

[1300] 9. The terminal displays to the user, "Product B is in stock."

[1301] Prompt Sentence Examples

[1302] "Please tell me the price of product A."

[1303] "Please let me know the stock status of product B."

[1304] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI models and sales representatives.

[1305] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1306] Step 1:

[1307] The user submits a query. The user enters a question into the terminal interface. For example, "What is the price of product A?" The terminal sends this question to the server as JSON format data. The input is the user's question, and the output is the JSON data sent to the server.

[1308] Step 2:

[1309] The server receives the question. The server receives the JSON data sent from the device, analyzes the data, and extracts the question content. For example, the information may include "Price of Product A." The input is the JSON data from the device, and the output is the extracted question content.

[1310] Step 3:

[1311] The server forwards the question to the generative AI model. The server generates an API request to pass the extracted question content to the generative AI model. The request includes the question content and the user's identification information. The input is the extracted question content and identification information, and the output is an API request to the generative AI model.

[1312] Step 4:

[1313] The generative AI model generates an answer. The generative AI model accesses an internal database to search for appropriate information for the question. For example, it obtains information such as "The price of product A is 10,000 yen" and generates an answer. The generated answer is returned to the server as an API response. The input is the API request, and the output is the generated answer.

[1314] Step 5:

[1315] The server receives and verifies the answer. The server receives the answer returned by the generative AI model and verifies its contents. For example, it uses internal logic to check the accuracy of the price information. The input is the generated answer, and the output is the verification result.

[1316] Step 6:

[1317] The server sets an escalation flag. If the server determines that the answer is inaccurate based on the validation results, it sets an escalation flag. For example, it sets a flag when the answer is ambiguous or requires additional confirmation. The input is the validation result, and the output is the escalation flag.

[1318] Step 7:

[1319] The server performs the escalation process. If the escalation flag is set, the server sends the generated response to the sales representative. This is done using an interface or email system dedicated to the sales representative. The input is the escalation flag and the generated response, and the output is the data to be sent to the sales representative.

[1320] Step 8:

[1321] The sales representative makes the final confirmation. The sales representative receives the escalated response via the interface and checks the contents. For example, they may correct the response to "The actual price of product A was 12,000 yen." The input is the escalated response, and the output is the corrected response.

[1322] Step 9:

[1323] The server receives the final answer. The server receives the revised answer from the sales representative and verifies its contents. The input is the revised answer, and the output is the confirmed final answer.

[1324] Step 10:

[1325] The user receives the final answer. The server sends the final answer to the user's device. The device displays to the user, "The price of product A is 12,000 yen." The input is the final answer, and the output is the information displayed to the user.

[1326] (Application example 1)

[1327] Next, a description will be given of Application 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."

[1328] Conventional customer support systems often have delayed responses to inquiries, which reduces customer satisfaction. Furthermore, in situations where immediate response is required in a physical store, it is difficult to provide a fast and accurate response. The present invention aims to solve these problems and provide a system that streamlines customer support in physical stores.

[1329] 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.

[1330] In this invention, the server includes means for receiving questions from users, means for generating answers using generative AI, means for verifying the accuracy of the answers, means for escalation, means for providing the answers to users after final confirmation by a sales representative, and means for displaying the answers on a terminal installed in a physical store. This makes it possible to provide quick and accurate answers to inquiries in a physical store and improve customer satisfaction.

[1331] "Means for receiving questions from users" is a function that allows the system to import questions entered by customers in a physical store using a tablet device, smartphone, etc.

[1332] "Generative AI" is an artificial intelligence model that automatically generates appropriate answers to questions from users, and often uses natural language processing technology.

[1333] "Means for verifying the accuracy of answers" is a function for checking whether the answers generated by generative AI are correct.

[1334] "Means for escalation" is a function that includes a process for raising the content of a generated answer to a human such as a sales representative if there are concerns about its accuracy.

[1335] The "means of providing the answer to the user after final confirmation by the sales representative" is a function for presenting the final answer that has been confirmed and corrected by the sales representative to the user.

[1336] "Means for displaying answers on a terminal installed in a physical store" is a function for displaying generated answers and corrected answers on a tablet terminal or display in a physical store.

[1337] This invention aims to improve the efficiency of customer support in brick-and-mortar stores. This system receives questions from users, uses generative AI to quickly and accurately generate answers, and verifies their appropriateness to improve customer satisfaction. The system is primarily composed of the following hardware and software:

[1338] System configuration and technologies used

[1339] 1. User Interface

[1340] The system uses tablets or smartphones installed in physical stores, allowing users to input specific questions.

[1341] 2. Server

[1342] It manages the core processing of the system, including receiving questions, generating answers using generative AI, verifying answers, escalating, final confirmation, and providing answers. It uses a general cloud server or on-premise server.

[1343] 3. Generative AI

[1344] It uses generative AI, such as OpenAI's API, to automatically generate appropriate answers to user questions, retrieving information from an internal database and generating prompts.

[1345] 4. Internal Database

[1346] We use a database that stores information necessary for sales activities, such as product information, price information, and inventory information.

[1347] 5. Escalation function

[1348] The server verifies the appropriateness of the generated answers and escalates to a sales representative if an incorrect answer is generated or requires clarification.

[1349] 6. Salesperson Interface

[1350] Provides an interface for sales representatives to review escalated content and make corrections or additions.

[1351] 7. Answer display function

[1352] The generated answers and the final answers confirmed and revised by the sales representative are displayed on tablet devices or displays in the physical store.

[1353] Processing flow

[1354] 1. User inputs a question

[1355] A user uses a tablet device in a physical store to type a question such as "Is this item in stock?"

[1356] 2. Receiving questions and generating answers using AI

[1357] The server receives the question and generates a prompt using the OpenAI API.

[1358] Example prompt: "User asks: Is this item in stock? Generate an appropriate answer."

[1359] 3. Response verification and escalation

[1360] The server verifies the generated answer, for example, "The item is not in stock. Please ask a store clerk for confirmation."

[1361] Escalate to sales if inaccurate or requires clarification.

[1362] 4. Final confirmation by sales representative

[1363] The sales representative checks the inventory status and corrects it to "This product is in stock."

[1364] 5. View Answers

[1365] The server receives the final answer and displays it on a tablet device in the physical store.

[1366] In this way, the present invention enables fast and accurate customer service in physical stores through collaboration between generative AI and sales representatives. Furthermore, the processing flow can be flexibly configured according to specific scenarios, aiming to improve customer satisfaction.

[1367] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1368] Step 1:

[1369] A user uses a tablet device in a physical store to enter a question. For example, they enter a question like "Is this product in stock?" The entered question is sent to the server by an application on the tablet device. The input data is in text format and is sent as an HTTP request to the server's question reception endpoint.

[1370] Step 2:

[1371] The server receives a question from a user. The received question is in text format and is first stored in a database. The server then generates a prompt to send the question to the generative AI. This prompt serves as an instruction for the generative AI model to generate an appropriate answer. For example, the format is "User question: Is this product in stock? Please generate an appropriate answer."

[1372] Step 3:

[1373] The generative AI receives the prompt sent from the server and generates an answer. The generative AI model (e.g., OpenAI GPT-3) generates an appropriate answer based on the prompt. This answer is generated using natural language processing technology and returned to the server in text format. An example of a generated answer is, "We cannot confirm the item is in stock. Please ask a store clerk to check."

[1374] Step 4:

[1375] The server verifies the accuracy of the answer returned by the generative AI. This verification is done by comparing it with information from an internal database. For example, if inventory information is not present in the database or if the AI's answer contains uncertainty, the server raises an escalation flag. If the answer is determined to be accurate, it proceeds to the next step.

[1376] Step 5:

[1377] If the server sets an escalation flag, it sends the question and the generated answer to the sales representative. The sending method is mainly via an internal API, and the sales representative's interface is notified. Here, the sales representative checks the question and the generated answer, and makes corrections or additions as necessary.

[1378] Step 6:

[1379] The sales representative receives the escalated information and checks the status. For example, they check the inventory information and enter a revised response into the system, such as "This product is in stock." The sales representative then sends the revised response to the server. This is mainly done via an internal API, and the server receives the revised response.

[1380] Step 7:

[1381] The server receives the final response from the sales representative and provides it to the user. The final response is sent to an application on a tablet device installed in the physical store and displayed to the user. The final response is then displayed on the user's screen, stating, "This product is in stock."

[1382] As described above, through the data processing and data calculation performed at each step, a system that provides quick and accurate answers to user inquiries is realized.

[1383] Furthermore, 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.

[1384] This invention is a system that provides prompt and accurate answers to inquiries from users in corporate sales activities. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the system aims to improve the user experience. Specifically, the emotion engine is used to detect the user's emotions and adjust the priority of responses and escalations based on those emotions.

[1385] System configuration

[1386] The system consists of the following main components:

[1387] 1. User Interface

[1388] The terminal (user device) provides an interface for the user to enter questions. The user interface can be provided in a variety of forms, such as a web or mobile application.

[1389] 2. Server

[1390] Manages the core processing of the system, including question receipt, answer generation, answer validation, escalation, final confirmation, answer delivery, and emotion recognition.

[1391] 3. Generative AI

[1392] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and constructing appropriate answers based on that information.

[1393] 4. Internal Database

[1394] This is an internal company database that is accessed by generative AI and contains product information, pricing information, inventory information, etc.

[1395] 5. Emotion Engine

[1396] This engine recognizes emotions from the user's input and writing style. For example, if the emotion engine detects negative emotions, it adjusts the responses generated by the generative AI and, if necessary, increases the priority of escalation.

[1397] 6. Escalation function

[1398] The server has a function of transmitting the inquiry and the generated answer to a sales representative when the user is unsure about the generated answer or when the user shows negative emotions.

[1399] 7. Salesperson Interface

[1400] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[1401] Program processing

[1402] 1. User inquiry submission

[1403] The user enters a question: "What is the price of product A?"

[1404] The terminal (user's device) sends this question to the server.

[1405] 2. Processing on the server

[1406] The server receives the user's query.

[1407] The server sends the question to the emotion engine to recognize the user's emotion.

[1408] 3. Generative AI Processing

[1409] The server sends the question to the generative AI.

[1410] Generative AI retrieves relevant information from an internal database and generates an appropriate answer to the question, for example, "The price of product A is 10,000 yen."

[1411] 4. Emotion Recognition by Emotion Engine

[1412] The emotion engine recognizes emotions from the user's input and writing style, for example, if the user expresses dissatisfaction or doubt.

[1413] 5. Response verification on the server

[1414] The server verifies the accuracy of the generated answer.

[1415] If the verification results show no problems, the answer is provided directly to the user.

[1416] 6. Response adjustment using emotion engine

[1417] The emotion engine adjusts the generative AI's responses based on the user's emotions, for example, using more polite language when negative emotions are recognized.

[1418] 7. Escalation Process

[1419] If the server is unsure about the generated answer or if the sentiment engine recognizes a negative sentiment, it raises an escalation flag.

[1420] The server sends the inquiry and generated response to the sales representative.

[1421] 8. Final confirmation by sales representative

[1422] Sales representatives receive escalated questions and answers and can review and correct them.

[1423] 9. Providing answers to users

[1424] After the server receives the final response from the sales representative, it provides the response to the user.

[1425] The terminal displays to the user, "The price of product A is 10,000 yen."

[1426] Specific example explanation

[1427] Example 1: When confirmation is not required

[1428] 1. A user asks, "What is the price of product A?"

[1429] 2. The device sends this question to the server.

[1430] 3. The server receives the question and sends it to the generative AI and emotion engine.

[1431] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[1432] 5. The emotion engine recognizes the user's emotions and does not detect negative emotions in particular.

[1433] 6. The server receives this response and determines that there is no problem with the verification.

[1434] 7. The server provides the user with the answer, "The price of product A is 10,000 yen."

[1435] 8. The device displays the answer to the user.

[1436] Example 2: When confirmation is required

[1437] 1. A user asks, "What is the stock status of product B?"

[1438] 2. The device sends this question to the server.

[1439] 3. The server receives the question and sends it to the generative AI and emotion engine.

[1440] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[1441] 5. The emotion engine recognizes when a user has anxiety.

[1442] 6. The server sets an escalation flag.

[1443] 7. The server sends the inquiry and generated response to the sales representative.

[1444] 8. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[1445] 9. The server provides the final answer to the user.

[1446] 10. The terminal displays to the user, "Product B is in stock."

[1447] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI, an emotion engine, and sales representatives, and also realizes flexible responses that correspond to the user's emotions.

[1448] The processing flow will be explained below.

[1449] Step 1:

[1450] The user enters a question: "What is the price of product A?"

[1451] Step 2:

[1452] The terminal sends a question to the server.

[1453] Step 3:

[1454] The server receives the user's query.

[1455] Step 4:

[1456] The server sends the question to the emotion engine.

[1457] Step 5:

[1458] The emotion engine analyzes the content of the question and recognizes the user's emotions, for example, assessing whether the user is anxious or suspicious based on the style and wording.

[1459] Step 6:

[1460] The server sends the question to the generative AI.

[1461] Step 7:

[1462] The generative AI analyzes the question and retrieves relevant information from an internal database.

[1463] Step 8:

[1464] Generative AI generates answers to questions, such as "The price of product A is 10,000 yen."

[1465] Step 9:

[1466] The answer generated by the generative AI is sent to the server.

[1467] Step 10:

[1468] The server receives the generated answer.

[1469] Step 11:

[1470] The server verifies the accuracy of the answer it receives, for example, by checking that it matches the latest information in its database.

[1471] Step 12:

[1472] The emotion engine adjusts responses based on the user's perceived emotions, for example adding the phrase "Don't worry" to the response if it detects that the user is anxious.

[1473] Step 13:

[1474] If the server determines based on the verification result that confirmation is not necessary, it sends the answer directly to the user.

[1475] Step 14:

[1476] The user's device receives and displays the answer. For example, it displays the answer "The price of product A is 10,000 yen."

[1477] Step 15:

[1478] If the server is unsure about the generated answer or if the sentiment engine detects a negative sentiment, it raises an escalation flag.

[1479] Step 16:

[1480] The server sends the response flagged for escalation to the sales representative.

[1481] Step 17:

[1482] Sales reps receive and review escalated questions and answers.

[1483] Step 18:

[1484] The sales representative checks the answer, corrects it if necessary, and sends the final answer to the server.

[1485] Step 19:

[1486] The server sends a final, confirmed response to the user, for example, "Product B is in stock."

[1487] Step 20:

[1488] The user's device receives and displays the final response, for example, "Product B is in stock."

[1489] The above is the flow of specific processing steps in a system that combines emotion engines.

[1490] Example 2

[1491] Next, a description will be given of 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."

[1492] In modern corporate sales activities, it is necessary to provide fast and accurate answers to user inquiries while also taking user emotions into consideration. However, existing systems lack accuracy and speed in their responses, and fail to properly reflect user emotions. This not only impairs the user experience, but also reduces sales efficiency.

[1493] 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.

[1494] In this invention, the server includes means for receiving questions from users, means for generating answers using generative AI, means for verifying the accuracy of the generated answers, emotion engine means for recognizing the user's emotions based on the content of the question, means for adjusting the content of the generated answers based on the emotions recognized by the emotion engine, means for escalating the question as necessary based on the results of the verification and emotion recognition, and means for providing the answer to the user after final confirmation by a sales representative. This enables the provision of fast and accurate answers and flexible responses based on the user's emotions.

[1495] "User" means a corporate or individual user who submits a question using the system.

[1496] A "means for receiving a query" is a component of the system for receiving and processing a query sent from a user's device.

[1497] "Generative AI" is an artificial intelligence engine that automatically generates answers to user questions based on an internal database.

[1498] "Means for generating answers" refers to the system's functionality for using generative AI to construct appropriate answers to users' questions.

[1499] "Means for verifying the accuracy of answers" refers to the system's process for checking that the answers generated contain appropriate and correct information.

[1500] The "emotion engine" is the engine of the system that recognizes emotions from the user's input content and writing style, and adjusts the response based on those emotions.

[1501] "Emotion recognition means" refers to the system's function of analyzing and evaluating the user's emotions through an emotion engine.

[1502] The "means for adjusting the content of the answer" is a function of the system that changes the wording and content of the generated answer based on the user's emotions recognized by the emotion engine.

[1503] "Means for escalation" refers to the system's function of passing on the inquiry and answer to a higher-level sales representative as needed, based on the generated answer and the user's sentiment.

[1504] "Sales Representative" means a company employee whose job is to provide final review of escalated inquiries and make corrections as necessary.

[1505] The "means of providing the answer to the user after final confirmation" is a system function that sends the final answer to the user after the sales representative has confirmed and corrected it.

[1506] This invention is a system for quickly and accurately responding to inquiries from users in corporate sales activities. This system receives questions from users, generates answers using generative AI, and uses an emotion engine to realize flexible responses that recognize and consider the user's emotions. Detailed embodiments of the present invention are described below.

[1507] System configuration

[1508] 1. User Interface

[1509] Users use a terminal (user device) to input questions using a web browser or mobile application. The user device provides an interface for sending the question data to the server.

[1510] 2. Server

[1511] The server manages the core processing of the system, specifically receiving questions, generating answers using generative AI, verifying the accuracy of the answers, recognizing emotions using an emotion engine, handling escalations, and providing answers to users.

[1512] 3. Generative AI

[1513] The generative AI works in conjunction with the server, retrieving relevant information from an internal database and automatically generating answers to user questions. The generative AI model uses natural language processing technology.

[1514] 4. Internal Database

[1515] Internal databases store necessary company information such as product information, price information, inventory information, etc. Generative AI accesses these databases and generates answers to questions.

[1516] 5. Emotion Engine

[1517] The emotion engine recognizes the emotion in a user's input and writing style, adjusts the content and wording of the response based on that emotion, and also detects when a user is expressing negative emotions and prioritizes escalation.

[1518] 6. Escalation function

[1519] If the server is concerned about the generated answer or if the emotion engine recognizes a negative emotion, it sets an escalation flag and sends the inquiry and the generated answer to the sales representative.

[1520] 7. Salesperson Interface

[1521] Sales representatives receive escalated content and have an interface for reviewing and correcting it, which consists of a web dashboard and a dedicated mobile application.

[1522] Specific examples

[1523] Example 1: No confirmation required

[1524] 1. A user asks, "What is the price of product A?"

[1525] 2. The device sends this question to the server.

[1526] 3. The server receives the question and sends it to the generative AI and emotion engine.

[1527] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[1528] 5. The emotion engine recognizes the user's emotions and does not detect negative emotions.

[1529] 6. The server receives this response and determines that there is no problem with the verification.

[1530] 7. The server provides the user with the answer, "The price of product A is 10,000 yen."

[1531] 8. The device displays the answer to the user.

[1532] Example 2: When confirmation is required

[1533] 1. A user asks, "What is the stock status of product B?"

[1534] 2. The device sends this question to the server.

[1535] 3. The server receives the question and sends it to the generative AI and emotion engine.

[1536] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[1537] 5. The emotion engine recognizes when a user is anxious.

[1538] 6. The server sets an escalation flag.

[1539] 7. The server sends the inquiry and generated response to the sales representative.

[1540] 8. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[1541] 9. The server provides the final answer to the user.

[1542] 10. The terminal displays to the user, "Product B is in stock."

[1543] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI, an emotion engine, and sales representatives, and also realizes flexible responses that correspond to the user's emotions.

[1544] Prompt Sentence Examples

[1545] "Please tell me the price of product A."

[1546] "Please let me know the stock status of product B."

[1547] "I would like to know the detailed specifications of product C."

[1548] By using the above, it is possible to respond appropriately and quickly to a wide variety of questions from users.

[1549] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1550] Step 1:

[1551] The user inputs a question such as "What is the price of product A?"

[1552] Specific actions: A user enters a question into an input form in a web browser or mobile application and presses the "Submit" button.

[1553] Input: Text input: "What is the price of product A?"

[1554] Output: The question is sent from the device to the server.

[1555] Step 2:

[1556] The terminal sends this question to the server.

[1557] Specific operation: The terminal sends the user's input to the server as an HTTP request.

[1558] Input: The question entered by the user.

[1559] Output: The question is sent to the server.

[1560] Step 3:

[1561] The server receives the user's query.

[1562] Specific operation: The server parses the HTTP request and extracts the question.

[1563] Input: The question sent from the device.

[1564] Output: The query is ready to be processed by the server.

[1565] Step 4:

[1566] The server sends the question content to the emotion engine and recognizes the emotion.

[1567] Specific operation: The server passes the question content to the emotion engine API and obtains the emotion score.

[1568] Input: The question.

[1569] Output: The emotion score returned by the emotion engine (e.g., happy, anger, sadness, etc.).

[1570] Step 5:

[1571] The server sends the question to the generative AI, which generates an answer.

[1572] How it works: The server calls the generative AI's API and inputs the question. The generative AI retrieves relevant information from its internal database and generates an appropriate answer.

[1573] Input: The question.

[1574] Output: The answer returned by the generative AI (e.g., "Product A costs 10,000 yen").

[1575] Step 6:

[1576] The server verifies the accuracy of the generated answer.

[1577] What happens: The server uses internal logic to check whether the generated answer is of the correct format and content, e.g., whether the price information is in the correct format.

[1578] Input: Answer from a generative AI.

[1579] Output: Answers verified for accuracy.

[1580] Step 7:

[1581] The server adjusts the response content based on the emotion recognized by the emotion engine.

[1582] Specific behavior: The server takes into account the emotion score from the emotion engine and adjusts the style and content of the response. For example, if a negative emotion is detected, it will change the wording to be more polite.

[1583] Input: Sentiment score, generated answer.

[1584] Output: The adjusted answer.

[1585] Step 8:

[1586] The server sets an escalation flag if necessary and sends the inquiry and generated response to a sales representative.

[1587] What it does: If the server is unsure about the generated answer or if the sentiment engine recognizes a negative sentiment, it sets an escalation flag and sends the inquiry and answer to a sales rep.

[1588] Input: A tailored response, and an escalation flag if necessary.

[1589] Output: Enquiry and response sent to sales representative.

[1590] Step 9:

[1591] Sales representatives receive escalated questions and answers and can review and correct them.

[1592] What happens next? Salespeople review details in a dedicated interface (web dashboard or mobile app) and enter revised responses as needed.

[1593] Input: Escalated question and answer.

[1594] Output: The final revised answer.

[1595] Step 10:

[1596] After receiving the final response from the sales representative, the server provides the response to the user.

[1597] Specific operation: The server sends the final answer to the terminal as an HTTP response.

[1598] Input: Final response from sales representative.

[1599] Output: The final answer sent to the user.

[1600] Step 11:

[1601] The terminal displays the final answer to the user.

[1602] Specific operation: The device receives the response from the server and displays the final answer in the user interface.

[1603] Input: The final response from the server.

[1604] Output: The final answer that is displayed to the user.

[1605] The above are the specific processing steps of this system, which allows for quick and accurate answers and flexible responses that reflect the user's emotions.

[1606] (Application example 2)

[1607] Next, a description will be given of Application 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."

[1608] Traditional corporate sales activities and customer support systems face challenges in providing prompt and accurate responses to user inquiries. Furthermore, a lack of response that reflects user emotions poses a risk of a poor user experience. In particular, if users with negative emotions are not properly addressed, customer satisfaction may decline, adversely affecting business growth. To solve these issues, a system that combines generative AI and emotion recognition technology and has an appropriate escalation function is needed.

[1609] 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.

[1610] In this invention, the server includes a means for receiving questions from users, a means for generating answers using a generative AI model, a means for using an emotion engine to recognize user emotions, a means for escalating if a negative emotion is recognized, and a means for providing an answer to the user after final confirmation by a sales representative. This enables quick and accurate responses to user inquiries and flexible responses according to emotions.

[1611] A "server" is a computer system that provides data and services in response to requests from other computers or devices.

[1612] "User" refers to any individual or legal entity that uses the system or services.

[1613] A "Question" is text or audio information provided by a User in the form of an inquiry or request for information.

[1614] A "generative AI model" is an artificial intelligence technology that automatically generates answers based on questions posed by users.

[1615] "Answer" refers to the information or solution provided by a generative AI model or a salesperson.

[1616] "Accuracy" refers to information or answers that are factual and free from errors.

[1617] An "emotion engine" is a technology that recognizes emotions from a user's questions or input, and provides appropriate responses and responses based on those emotions.

[1618] "Negative affect" refers to negative emotional states such as anxiety, doubt, anger, and sadness.

[1619] "Escalation" is the process of passing a difficult problem or situation to a higher level person or expert.

[1620] "Salesperson" is the staff member responsible for communicating with the user and confirming or correcting the final response.

[1621] An "internal database" is a data storage system installed within a company that a generative AI model accesses and retrieves information from.

[1622] This invention is a system for quickly and accurately responding to inquiries in corporate sales activities and customer support systems. In particular, it aims to improve the user experience by identifying the user's emotions and providing answers or escalations based on those emotions. The specific configuration and operation of this system are described in detail below.

[1623] System configuration

[1624] The system consists of the following main components:

[1625] 1. Server

[1626] The server receives questions from users, generates answers using a generative AI model, recognizes user emotions using an emotion engine, and escalates the situation as needed. The server also provides the final answer.

[1627] 2. Generative AI Models

[1628] The generative AI model retrieves information from an internal database and generates appropriate answers based on user questions. The generative AI model is built using natural language processing technology.

[1629] 3. Emotion Engine

[1630] The emotion engine recognizes emotions from user questions and inputs, especially negative emotions (e.g., anxiety or doubt), and adjusts the answers of the generative AI model.

[1631] 4. Escalation function

[1632] If the emotion engine recognizes a negative emotion or if the generated response is unsatisfactory, the server automatically escalates the inquiry and sends the generated response to a sales representative.

[1633] 5. Salesperson Interface

[1634] The sales representative receives the escalated questions and answers, checks and corrects them, and the final corrected answers are provided to the user.

[1635] Hardware and software used

[1636] Smartphone: Used to accept input from users and display results.

[1637] Server: Receives questions, generates answers, recognizes emotions, and handles escalation.

[1638] Generative AI model: Obtains information from an internal database and generates answers.

[1639] Emotion engine: Recognizes user emotions and tailors responses.

[1640] Specific examples

[1641] 1. Receiving user questions

[1642] A user uses a smartphone app to ask, "Can I return product X?" This question is sent to the server.

[1643] 2. Generating answers to questions

[1644] The generative AI model generates the answer, "Returns of product X are generally possible. Click here for details."

[1645] 3. User Emotion Recognition

[1646] The emotion engine recognizes the emotion of doubt from the user's question.

[1647] 4. Escalation

[1648] The server escalates based on the results of the emotion engine and sends the question and AI answer to the sales representative.

[1649] 5. Final confirmation and response

[1650] The sales representative will check the request and provide the final answer to the user via the server: "Returns are possible. If you have any questions, please contact our support center." The answer will be displayed on the smartphone.

[1651] Prompt Sentence Examples

[1652] Prompt for the AI ​​model to perform sentiment analysis:

[1653] Please analyze the following text and classify the emotion: 'Can I return product X?'

[1654] Prompt for generating an answer:

[1655] Generate a customer support response for the following question: 'Can I return product X?' using the company database.

[1656] In this way, the present invention integrates generative AI, an emotion engine, and an escalation function to realize an inquiry response system that can respond flexibly and appropriately according to the user's emotions.

[1657] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1658] Step 1:

[1659] The user inputs and sends a question using a device (smartphone app). An example of a user input is, "Can I return product X?" This input is sent to the server.

[1660] Step 2:

[1661] The server receives the question from the user, which is received in text format and stored internally for further processing.

[1662] Step 3:

[1663] The server sends the received question to the emotion engine, which recognizes the user's emotion from the question. For example, the emotion engine recognizes doubt from the question "Are returns possible?" The processing result is returned to the server.

[1664] Step 4:

[1665] The server sends the question and the emotion data obtained from the emotion engine to the generative AI model. The generative AI model retrieves the corresponding information from its internal database and generates an answer to the question. For example, it generates an answer such as, "Returns of product X are usually possible. Click here for details." This answer is then returned to the server.

[1666] Step 5:

[1667] The server verifies the answer obtained from the generative AI model. During this verification process, data from an internal database is referenced to confirm the accuracy and appropriateness of the answer. If there are no concerns as a result of the verification, the server proceeds to the next step.

[1668] Step 6:

[1669] The server uses data from the emotion engine to determine if the user has a negative emotion, and if a negative emotion is recognized, the server sets an escalation flag.

[1670] Step 7:

[1671] If a negative sentiment is detected, the server automatically escalates the query and sends the generated response to a sales representative.

[1672] Step 8:

[1673] The sales representative checks the received question and the AI-generated answer and modifies it as necessary. For example, if the answer is modified to "Returns are possible. If you have any questions, please contact our support center," the modified answer is returned to the server.

[1674] Step 9:

[1675] The server receives the final confirmed answer from the sales representative and provides it to the user. This answer is displayed on the device (smartphone app) and notified to the user.

[1676] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio 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 audio data.

[1677] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with 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.

[1678] 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 headset type terminal 314.

[1679] [Fourth embodiment]

[1680] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1681] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1682] 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 WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1683] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. 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 control target 443 are also connected to the bus 52.

[1684] 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 audio in accordance with instructions from the processor 46.

[1685] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and 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 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1686] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1687] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1688] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, 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.

[1689] 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.

[1690] 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.

[1691] In the robot 414, the processor 46 performs the reception output process. 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.

[1692] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1693] This invention is a system that provides fast and accurate answers to inquiries from users in sales activities for corporations. Specifically, it uses generative AI to automatically generate answers, verifies their accuracy, and escalates them to a sales representative as necessary, thereby achieving both speed and quality in responses to users.

[1694] System configuration

[1695] The system consists of the following main components:

[1696] 1. User Interface

[1697] The terminal (user device) provides an interface for the user to enter questions. The user interface can be provided in a variety of forms, such as a web or mobile application.

[1698] 2. Server

[1699] Manages the core processes of the system, including receiving questions, generating answers, verifying answers, escalating, finalizing, and providing answers.

[1700] 3. Generative AI

[1701] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and constructing appropriate answers based on that information.

[1702] 4. Internal Database

[1703] This is an internal company database that is accessed by generative AI and contains product information, pricing information, inventory information, etc.

[1704] 5. Escalation function

[1705] The server has a function of transmitting the inquiry and the generated response to a sales representative when there is any concern about the generated response or when confirmation of accuracy is required.

[1706] 6. Salesperson Interface

[1707] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[1708] Program processing

[1709] 1. User inquiry submission

[1710] The user enters a question: "What is the price of product A?"

[1711] The terminal (user's device) sends this question to the server.

[1712] 2. Processing on the server

[1713] The server receives the query sent from the terminal.

[1714] The server sends the received questions to the generative AI.

[1715] 3. Generative AI Processing

[1716] Generative AI generates appropriate answers to questions.

[1717] For example, the answer "The price of product A is 10,000 yen" is retrieved from an internal database and generated.

[1718] The generated answer is sent back to the server.

[1719] 4. Response verification on the server

[1720] The server verifies the accuracy of the generated answer.

[1721] If the verification results show no problems, the answer is provided directly to the user.

[1722] If you are unsure about the content or need clarification, raise an escalation flag.

[1723] 5. Escalation Process

[1724] If the escalation flag is set, the server sends the inquiry and generated response to the sales representative.

[1725] 6. Final confirmation by sales representative

[1726] The sales representative receives the escalated content and checks and corrects it.

[1727] The final corrected answer is sent to the server.

[1728] 7. Providing answers to users

[1729] After the server receives the final response from the sales representative, it provides the response to the user.

[1730] The terminal displays to the user, "The price of product A is 10,000 yen."

[1731] Specific example explanation

[1732] Example 1: When confirmation is not required

[1733] 1. A user asks, "What is the price of product A?"

[1734] 2. The device sends this question to the server.

[1735] 3. The server receives the question and sends it to the generative AI.

[1736] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[1737] 5. The server receives this response and determines that there is no problem with the verification.

[1738] 6. The server provides the user with the answer, "The price of product A is 10,000 yen."

[1739] 7. The terminal displays the answer to the user.

[1740] Example 2: When confirmation is required

[1741] 1. A user asks, "What is the stock status of product B?"

[1742] 2. The device sends this question to the server.

[1743] 3. The server receives the question and sends it to the generative AI.

[1744] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[1745] 5. The server receives this response and sets the escalation flag.

[1746] 6. The server sends the inquiry and generated response to the sales representative.

[1747] 7. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[1748] 8. The server receives the final response and provides it to the user.

[1749] 9. The terminal displays to the user, "Product B is in stock."

[1750] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI and sales representatives.

[1751] The processing flow will be explained below.

[1752] Step 1:

[1753] The user enters a question, for example, "What is the price of product A?"

[1754] Step 2:

[1755] The terminal sends the user's question to the server.

[1756] Step 3:

[1757] The server receives the user's query.

[1758] Step 4:

[1759] The server sends the question to the generative AI.

[1760] Step 5:

[1761] The generative AI receives the question and retrieves relevant information from an internal database.

[1762] Step 6:

[1763] Generative AI generates an appropriate answer to the question, for example, "The price of product A is 10,000 yen."

[1764] Step 7:

[1765] The answer generated by the generative AI is sent to the server.

[1766] Step 8:

[1767] The server receives the answer from the generative AI.

[1768] Step 9:

[1769] The server verifies the accuracy of the generated answer.

[1770] Step 10:

[1771] If the server determines that confirmation is not necessary based on the verification results, it sends the answer to the user as is. For example, it sends the answer "The price of product A is 10,000 yen."

[1772] Step 11:

[1773] The user's terminal receives and displays the response.

[1774] Step 12:

[1775] If the server is unsure about the generated answer, it will raise an escalation flag.

[1776] Step 13:

[1777] The server sends the response flagged for escalation to the sales representative.

[1778] Step 14:

[1779] Sales reps receive and review escalated questions and answers.

[1780] Step 15:

[1781] After the sales representative checks the answer, he / she corrects the answer as necessary and sends the final answer to the server.

[1782] Step 16:

[1783] The server sends a final, confirmed response to the user, for example, "Product B is in stock."

[1784] Step 17:

[1785] The user's device receives and displays the final response, for example, "Product B is in stock."

[1786] The above is a specific flow of processing steps in this system.

[1787] Example 1

[1788] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1789] In conventional corporate sales activities, it is difficult to provide quick and accurate responses to user inquiries, and there is a risk of delays in appropriate responses, especially when the inquiry is complex or requires accuracy. Furthermore, when human confirmation is required, the work of the person in charge increases and efficiency decreases, which raises concerns about lower customer satisfaction and a decline in the efficiency of sales activities. These issues need to be resolved.

[1790] 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.

[1791] In this invention, the server includes a means for receiving inquiries from users, a means for generating answers using a generative AI model, and a means for verifying the accuracy of the generated answers. This enables the automatic generation and provision of quick and accurate answers. Furthermore, if there are concerns about the accuracy of the answer, an escalation flag is set and the inquiry and the generated answer are sent to a sales representative, who then provides an appropriate answer to the user after final confirmation. This makes it possible to achieve both speed and quality in answers, thereby improving customer satisfaction and streamlining sales activities.

[1792] "User" means any person or entity that queries the System.

[1793] "Inquiry" refers to the act of a user providing information or asking a question to the system.

[1794] A "generative AI model" refers to an artificial intelligence program that uses natural language processing technology to automatically generate answers to user inquiries.

[1795] "Answer" refers to the information or response generated by a generative AI model in response to a user's inquiry.

[1796] "Accuracy" refers to the degree to which the generated answer is based on facts and correct information.

[1797] "Verification" refers to the process of checking whether the generated answers are accurate.

[1798] "Escalation Flag" refers to an identifying signal within the system that is set when there is concern about the accuracy of the answer generated.

[1799] "Salesperson" refers to the person who makes the final confirmation when an escalation flag is raised.

[1800] "Final review" refers to the process in which a sales representative reviews the answers provided by the generative AI model and makes corrections if necessary.

[1801] "Internal database" refers to a database that stores information that a generative AI model accesses to generate answers.

[1802] "Escalation" refers to the process of seeking confirmation from a sales representative on a generated response.

[1803] "Automatic" refers to the way the system operates autonomously, without the need for human intervention.

[1804] "System" refers to a set of components with a series of functions that accept user inquiries, generate answers using generative AI models, verify accuracy, escalate as necessary, and ultimately provide answers to users.

[1805] This invention is a system that provides fast and accurate answers to inquiries from users in corporate sales activities. Specifically, the system automatically generates answers using a generative AI model, verifies their accuracy, and escalates the inquiries to sales representatives as necessary, thereby achieving both speed and quality in responses to users.

[1806] System configuration

[1807] The system consists of the following main components:

[1808] 1. User Interface

[1809] The user's device provides an interface for the user to input questions. The user interface can be provided in a variety of forms, such as a web browser or a mobile application.

[1810] 2. Server

[1811] Manages the core processes of the system, including receiving questions, generating answers, verifying answers, escalating, finalizing, and providing answers.

[1812] 3. Generative AI Models

[1813] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and building appropriate answers based on that information.

[1814] 4. Internal Database

[1815] An internal company database accessed by generative AI models, containing product information, pricing information, inventory information, etc.

[1816] 5. Escalation function

[1817] The server has a function of transmitting the inquiry and the generated response to a sales representative when there is any concern about the generated response or when confirmation of accuracy is required.

[1818] 6. Salesperson Interface

[1819] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[1820] Program processing flow

[1821] The process flow for realizing this system is explained below:

[1822] 1. User inquiry submission

[1823] The user types a question into the terminal interface, such as "What is the price of product A?", and the terminal sends this question to the server in JSON format or another suitable data format.

[1824] 2. Receiving and analyzing questions

[1825] The server receives the inquiry sent from the terminal, analyzes the input data, and extracts the question, for example, "What is the price of product A?"

[1826] 3. Answer generation using generative AI models

[1827] The server sends the extracted question to the generative AI model, which then accesses its internal database and generates the optimal answer to the question. For example, it generates an answer such as "The price of product A is 10,000 yen."

[1828] 4. Response verification and escalation

[1829] The server uses its internal logic to verify the accuracy of the generated answer, and if it determines that the answer is inaccurate, it sets an escalation flag. Based on the escalation flag, the server sends the question and the generated answer to the sales representative.

[1830] 5. Final confirmation by sales representative

[1831] The sales representative receives the escalated content on the sales representative interface, checks the accuracy of the response, corrects the response as necessary, and sends the final, accurate information to the server.

[1832] 6. Providing a final response

[1833] The server receives the final response from the sales representative and provides it to the user. The user's terminal displays this final response, such as "The price of product A is 12,000 yen."

[1834] Specific example explanation

[1835] Example 1: No confirmation required

[1836] 1. A user asks, "What is the price of product A?"

[1837] 2. The device sends this question to the server.

[1838] 3. The server receives the question and sends it to the generative AI model.

[1839] 4. The generative AI model generates the answer, "The price of Product A is 10,000 yen."

[1840] 5. The server receives this response and determines that there is no problem with the verification.

[1841] 6. The server provides the user with the answer, "The price of product A is 10,000 yen."

[1842] 7. The device displays the answer to the user.

[1843] Example 2: When confirmation is required

[1844] 1. A user asks, "What is the availability of product B?"

[1845] 2. The device sends this question to the server.

[1846] 3. The server receives the question and sends it to the generative AI model.

[1847] 4. The generative AI model generates the answer, "The stock of product B is unknown. You may need to check."

[1848] 5. The server receives this response and sets the escalation flag.

[1849] 6. The server sends the inquiry and generated response to the sales representative.

[1850] 7. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[1851] 8. The server receives the final response and provides it to the user.

[1852] 9. The terminal displays to the user, "Product B is in stock."

[1853] Prompt Sentence Examples

[1854] "Please tell me the price of product A."

[1855] "Please let me know the stock status of product B."

[1856] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI models and sales representatives.

[1857] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1858] Step 1:

[1859] The user submits a query. The user enters a question into the terminal interface. For example, "What is the price of product A?" The terminal sends this question to the server as JSON format data. The input is the user's question, and the output is the JSON data sent to the server.

[1860] Step 2:

[1861] The server receives the question. The server receives the JSON data sent from the device, analyzes the data, and extracts the question content. For example, the information may include "Price of Product A." The input is the JSON data from the device, and the output is the extracted question content.

[1862] Step 3:

[1863] The server forwards the question to the generative AI model. The server generates an API request to pass the extracted question content to the generative AI model. The request includes the question content and the user's identification information. The input is the extracted question content and identification information, and the output is an API request to the generative AI model.

[1864] Step 4:

[1865] The generative AI model generates an answer. The generative AI model accesses an internal database to search for appropriate information for the question. For example, it obtains information such as "The price of product A is 10,000 yen" and generates an answer. The generated answer is returned to the server as an API response. The input is the API request, and the output is the generated answer.

[1866] Step 5:

[1867] The server receives and verifies the answer. The server receives the answer returned by the generative AI model and verifies its contents. For example, it uses internal logic to check the accuracy of the price information. The input is the generated answer, and the output is the verification result.

[1868] Step 6:

[1869] The server sets an escalation flag. If the server determines that the answer is inaccurate based on the validation results, it sets an escalation flag. For example, it sets a flag when the answer is ambiguous or requires additional confirmation. The input is the validation result, and the output is the escalation flag.

[1870] Step 7:

[1871] The server performs the escalation process. If the escalation flag is set, the server sends the generated response to the sales representative. This is done using an interface or email system dedicated to the sales representative. The input is the escalation flag and the generated response, and the output is the data to be sent to the sales representative.

[1872] Step 8:

[1873] The sales representative makes the final confirmation. The sales representative receives the escalated response via the interface and checks the contents. For example, they may correct the response to "The actual price of product A was 12,000 yen." The input is the escalated response, and the output is the corrected response.

[1874] Step 9:

[1875] The server receives the final answer. The server receives the revised answer from the sales representative and verifies its contents. The input is the revised answer, and the output is the confirmed final answer.

[1876] Step 10:

[1877] The user receives the final answer. The server sends the final answer to the user's device. The device displays to the user, "The price of product A is 12,000 yen." The input is the final answer, and the output is the information displayed to the user.

[1878] (Application example 1)

[1879] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1880] Conventional customer support systems often have delayed responses to inquiries, which reduces customer satisfaction. Furthermore, in situations where immediate response is required in a physical store, it is difficult to provide a fast and accurate response. The present invention aims to solve these problems and provide a system that streamlines customer support in physical stores.

[1881] 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.

[1882] In this invention, the server includes means for receiving questions from users, means for generating answers using generative AI, means for verifying the accuracy of the answers, means for escalation, means for providing the answers to users after final confirmation by a sales representative, and means for displaying the answers on a terminal installed in a physical store. This makes it possible to provide quick and accurate answers to inquiries in a physical store and improve customer satisfaction.

[1883] "Means for receiving questions from users" is a function that allows the system to import questions entered by customers in a physical store using a tablet device, smartphone, etc.

[1884] "Generative AI" is an artificial intelligence model that automatically generates appropriate answers to questions from users, and often uses natural language processing technology.

[1885] "Means for verifying the accuracy of answers" is a function for checking whether the answers generated by generative AI are correct.

[1886] "Means for escalation" is a function that includes a process for raising the content of a generated answer to a human such as a sales representative if there are concerns about its accuracy.

[1887] The "means of providing the answer to the user after final confirmation by the sales representative" is a function for presenting the final answer that has been confirmed and corrected by the sales representative to the user.

[1888] "Means for displaying answers on a terminal installed in a physical store" is a function for displaying generated answers and corrected answers on a tablet terminal or display in a physical store.

[1889] This invention aims to improve the efficiency of customer support in brick-and-mortar stores. This system receives questions from users, uses generative AI to quickly and accurately generate answers, and verifies their appropriateness to improve customer satisfaction. The system is primarily composed of the following hardware and software:

[1890] System configuration and technologies used

[1891] 1. User Interface

[1892] The system uses tablets or smartphones installed in physical stores, allowing users to input specific questions.

[1893] 2. Server

[1894] It manages the core processing of the system, including receiving questions, generating answers using generative AI, verifying answers, escalating, final confirmation, and providing answers. It uses a general cloud server or on-premise server.

[1895] 3. Generative AI

[1896] It uses generative AI, such as OpenAI's API, to automatically generate appropriate answers to user questions, retrieving information from an internal database and generating prompts.

[1897] 4. Internal Database

[1898] We use a database that stores information necessary for sales activities, such as product information, price information, and inventory information.

[1899] 5. Escalation function

[1900] The server verifies the appropriateness of the generated answers and escalates to a sales representative if an incorrect answer is generated or requires clarification.

[1901] 6. Salesperson Interface

[1902] Provides an interface for sales representatives to review escalated content and make corrections or additions.

[1903] 7. Answer display function

[1904] The generated answers and the final answers confirmed and revised by the sales representative are displayed on tablet devices or displays in the physical store.

[1905] Processing flow

[1906] 1. User inputs a question

[1907] A user uses a tablet device in a physical store to type a question such as "Is this item in stock?"

[1908] 2. Receiving questions and generating answers using AI

[1909] The server receives the question and generates a prompt using the OpenAI API.

[1910] Example prompt: "User asks: Is this item in stock? Generate an appropriate answer."

[1911] 3. Response verification and escalation

[1912] The server verifies the generated answer, for example, "The item is not in stock. Please ask a store clerk for confirmation."

[1913] Escalate to sales if inaccurate or requires clarification.

[1914] 4. Final confirmation by sales representative

[1915] The sales representative checks the inventory status and corrects it to "This product is in stock."

[1916] 5. View Answers

[1917] The server receives the final answer and displays it on a tablet device in the physical store.

[1918] In this way, the present invention enables fast and accurate customer service in physical stores through collaboration between generative AI and sales representatives. Furthermore, the processing flow can be flexibly configured according to specific scenarios, aiming to improve customer satisfaction.

[1919] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1920] Step 1:

[1921] A user uses a tablet device in a physical store to enter a question. For example, they enter a question like "Is this product in stock?" The entered question is sent to the server by an application on the tablet device. The input data is in text format and is sent as an HTTP request to the server's question reception endpoint.

[1922] Step 2:

[1923] The server receives a question from a user. The received question is in text format and is first stored in a database. The server then generates a prompt to send the question to the generative AI. This prompt serves as an instruction for the generative AI model to generate an appropriate answer. For example, the format is "User question: Is this product in stock? Please generate an appropriate answer."

[1924] Step 3:

[1925] The generative AI receives the prompt sent from the server and generates an answer. The generative AI model (e.g., OpenAI GPT-3) generates an appropriate answer based on the prompt. This answer is generated using natural language processing technology and returned to the server in text format. An example of a generated answer is, "We cannot confirm the item is in stock. Please ask a store clerk to check."

[1926] Step 4:

[1927] The server verifies the accuracy of the answer returned by the generative AI. This verification is done by comparing it with information from an internal database. For example, if inventory information is not present in the database or if the AI's answer contains uncertainty, the server raises an escalation flag. If the answer is determined to be accurate, it proceeds to the next step.

[1928] Step 5:

[1929] If the server sets an escalation flag, it sends the question and the generated answer to the sales representative. The sending method is mainly via an internal API, and the sales representative's interface is notified. Here, the sales representative checks the question and the generated answer, and makes corrections or additions as necessary.

[1930] Step 6:

[1931] The sales representative receives the escalated information and checks the status. For example, they check the inventory information and enter a revised response into the system, such as "This product is in stock." The sales representative then sends the revised response to the server. This is mainly done via an internal API, and the server receives the revised response.

[1932] Step 7:

[1933] The server receives the final response from the sales representative and provides it to the user. The final response is sent to an application on a tablet device installed in the physical store and displayed to the user. The final response is then displayed on the user's screen, stating, "This product is in stock."

[1934] As described above, through the data processing and data calculation performed at each step, a system that provides quick and accurate answers to user inquiries is realized.

[1935] Furthermore, 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.

[1936] This invention is a system that provides prompt and accurate answers to inquiries from users in corporate sales activities. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the system aims to improve the user experience. Specifically, the emotion engine is used to detect the user's emotions and adjust the priority of responses and escalations based on those emotions.

[1937] System configuration

[1938] The system consists of the following main components:

[1939] 1. User Interface

[1940] The terminal (user device) provides an interface for the user to enter questions. The user interface can be provided in a variety of forms, such as a web or mobile application.

[1941] 2. Server

[1942] Manages the core processing of the system, including question receipt, answer generation, answer validation, escalation, final confirmation, answer delivery, and emotion recognition.

[1943] 3. Generative AI

[1944] It works in conjunction with the server to automatically generate answers to user questions, retrieving information from an internal database and constructing appropriate answers based on that information.

[1945] 4. Internal Database

[1946] This is an internal company database that is accessed by generative AI and contains product information, pricing information, inventory information, etc.

[1947] 5. Emotion Engine

[1948] This engine recognizes emotions from the user's input and writing style. For example, if the emotion engine detects negative emotions, it adjusts the responses generated by the generative AI and, if necessary, increases the priority of escalation.

[1949] 6. Escalation function

[1950] The server has a function of transmitting the inquiry and the generated answer to a sales representative when the user is unsure about the generated answer or when the user shows negative emotions.

[1951] 7. Salesperson Interface

[1952] Provide an interface for sales reps to receive and review escalations and make corrections or additions as needed.

[1953] Program processing

[1954] 1. User inquiry submission

[1955] The user enters a question: "What is the price of product A?"

[1956] The terminal (user's device) sends this question to the server.

[1957] 2. Processing on the server

[1958] The server receives the user's query.

[1959] The server sends the question to the emotion engine to recognize the user's emotion.

[1960] 3. Generative AI Processing

[1961] The server sends the question to the generative AI.

[1962] Generative AI retrieves relevant information from an internal database and generates an appropriate answer to the question, for example, "The price of product A is 10,000 yen."

[1963] 4. Emotion Recognition by Emotion Engine

[1964] The emotion engine recognizes emotions from the user's input and writing style, for example, if the user expresses dissatisfaction or doubt.

[1965] 5. Response verification on the server

[1966] The server verifies the accuracy of the generated answer.

[1967] If the verification results show no problems, the answer is provided directly to the user.

[1968] 6. Response adjustment using emotion engine

[1969] The emotion engine adjusts the generative AI's responses based on the user's emotions, for example, using more polite language when negative emotions are recognized.

[1970] 7. Escalation Process

[1971] If the server is unsure about the generated answer or if the sentiment engine recognizes a negative sentiment, it raises an escalation flag.

[1972] The server sends the inquiry and generated response to the sales representative.

[1973] 8. Final confirmation by sales representative

[1974] Sales representatives receive escalated questions and answers and can review and correct them.

[1975] 9. Providing answers to users

[1976] After the server receives the final response from the sales representative, it provides the response to the user.

[1977] The terminal displays to the user, "The price of product A is 10,000 yen."

[1978] Specific example explanation

[1979] Example 1: When confirmation is not required

[1980] 1. A user asks, "What is the price of product A?"

[1981] 2. The device sends this question to the server.

[1982] 3. The server receives the question and sends it to the generative AI and emotion engine.

[1983] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[1984] 5. The emotion engine recognizes the user's emotions and does not detect negative emotions in particular.

[1985] 6. The server receives this response and determines that there is no problem with the verification.

[1986] 7. The server provides the user with the answer, "The price of product A is 10,000 yen."

[1987] 8. The device displays the answer to the user.

[1988] Example 2: When confirmation is required

[1989] 1. A user asks, "What is the stock status of product B?"

[1990] 2. The device sends this question to the server.

[1991] 3. The server receives the question and sends it to the generative AI and emotion engine.

[1992] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[1993] 5. The emotion engine recognizes when a user has anxiety.

[1994] 6. The server sets an escalation flag.

[1995] 7. The server sends the inquiry and generated response to the sales representative.

[1996] 8. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[1997] 9. The server provides the final answer to the user.

[1998] 10. The terminal displays to the user, "Product B is in stock."

[1999] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI, an emotion engine, and sales representatives, and also realizes flexible responses that correspond to the user's emotions.

[2000] The processing flow will be explained below.

[2001] Step 1:

[2002] The user enters a question: "What is the price of product A?"

[2003] Step 2:

[2004] The terminal sends a question to the server.

[2005] Step 3:

[2006] The server receives the user's query.

[2007] Step 4:

[2008] The server sends the question to the emotion engine.

[2009] Step 5:

[2010] The emotion engine analyzes the content of the question and recognizes the user's emotions, for example, assessing whether the user is anxious or suspicious based on the style and wording.

[2011] Step 6:

[2012] The server sends the question to the generative AI.

[2013] Step 7:

[2014] The generative AI analyzes the question and retrieves relevant information from an internal database.

[2015] Step 8:

[2016] Generative AI generates answers to questions, such as "The price of product A is 10,000 yen."

[2017] Step 9:

[2018] The answer generated by the generative AI is sent to the server.

[2019] Step 10:

[2020] The server receives the generated answer.

[2021] Step 11:

[2022] The server verifies the accuracy of the answer it receives, for example, by checking that it matches the latest information in its database.

[2023] Step 12:

[2024] The emotion engine adjusts responses based on the user's perceived emotions, for example adding the phrase "Don't worry" to the response if it detects that the user is anxious.

[2025] Step 13:

[2026] If the server determines based on the verification result that confirmation is not necessary, it sends the answer directly to the user.

[2027] Step 14:

[2028] The user's device receives and displays the answer. For example, it displays the answer "The price of product A is 10,000 yen."

[2029] Step 15:

[2030] If the server is unsure about the generated answer or if the sentiment engine detects a negative sentiment, it raises an escalation flag.

[2031] Step 16:

[2032] The server sends the response flagged for escalation to the sales representative.

[2033] Step 17:

[2034] Sales reps receive and review escalated questions and answers.

[2035] Step 18:

[2036] The sales representative checks the answer, corrects it if necessary, and sends the final answer to the server.

[2037] Step 19:

[2038] The server sends a final, confirmed response to the user, for example, "Product B is in stock."

[2039] Step 20:

[2040] The user's device receives and displays the final response, for example, "Product B is in stock."

[2041] The above is the flow of specific processing steps in a system that combines emotion engines.

[2042] Example 2

[2043] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2044] In modern corporate sales activities, it is necessary to provide fast and accurate answers to user inquiries while also taking user emotions into consideration. However, existing systems lack accuracy and speed in their responses, and fail to properly reflect user emotions. This not only impairs the user experience, but also reduces sales efficiency.

[2045] 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.

[2046] In this invention, the server includes means for receiving questions from users, means for generating answers using generative AI, means for verifying the accuracy of the generated answers, emotion engine means for recognizing the user's emotions based on the content of the question, means for adjusting the content of the generated answers based on the emotions recognized by the emotion engine, means for escalating the question as necessary based on the results of the verification and emotion recognition, and means for providing the answer to the user after final confirmation by a sales representative. This enables the provision of fast and accurate answers and flexible responses based on the user's emotions.

[2047] "User" means a corporate or individual user who submits a question using the system.

[2048] A "means for receiving a query" is a component of the system for receiving and processing a query sent from a user's device.

[2049] "Generative AI" is an artificial intelligence engine that automatically generates answers to user questions based on an internal database.

[2050] "Means for generating answers" refers to the system's functionality for using generative AI to construct appropriate answers to users' questions.

[2051] "Means for verifying the accuracy of answers" refers to the system's process for checking that the answers generated contain appropriate and correct information.

[2052] The "emotion engine" is the engine of the system that recognizes emotions from the user's input content and writing style, and adjusts the response based on those emotions.

[2053] "Emotion recognition means" refers to the system's function of analyzing and evaluating the user's emotions through an emotion engine.

[2054] The "means for adjusting the content of the answer" is a function of the system that changes the wording and content of the generated answer based on the user's emotions recognized by the emotion engine.

[2055] "Means for escalation" refers to the system's function of passing on the inquiry and answer to a higher-level sales representative as needed, based on the generated answer and the user's sentiment.

[2056] "Sales Representative" means a company employee whose job is to provide final review of escalated inquiries and make corrections as necessary.

[2057] The "means of providing the answer to the user after final confirmation" is a system function that sends the final answer to the user after the sales representative has confirmed and corrected it.

[2058] This invention is a system for quickly and accurately responding to inquiries from users in corporate sales activities. This system receives questions from users, generates answers using generative AI, and uses an emotion engine to realize flexible responses that recognize and consider the user's emotions. Detailed embodiments of the present invention are described below.

[2059] System configuration

[2060] 1. User Interface

[2061] Users use a terminal (user device) to input questions using a web browser or mobile application. The user device provides an interface for sending the question data to the server.

[2062] 2. Server

[2063] The server manages the core processing of the system, specifically receiving questions, generating answers using generative AI, verifying the accuracy of the answers, recognizing emotions using an emotion engine, handling escalations, and providing answers to users.

[2064] 3. Generative AI

[2065] The generative AI works in conjunction with the server, retrieving relevant information from an internal database and automatically generating answers to user questions. The generative AI model uses natural language processing technology.

[2066] 4. Internal Database

[2067] Internal databases store necessary company information such as product information, price information, inventory information, etc. Generative AI accesses these databases and generates answers to questions.

[2068] 5. Emotion Engine

[2069] The emotion engine recognizes the emotion in a user's input and writing style, adjusts the content and wording of the response based on that emotion, and also detects when a user is expressing negative emotions and prioritizes escalation.

[2070] 6. Escalation function

[2071] If the server is concerned about the generated answer or if the emotion engine recognizes a negative emotion, it sets an escalation flag and sends the inquiry and the generated answer to the sales representative.

[2072] 7. Salesperson Interface

[2073] Sales representatives receive escalated content and have an interface for reviewing and correcting it, which consists of a web dashboard and a dedicated mobile application.

[2074] Specific examples

[2075] Example 1: No confirmation required

[2076] 1. A user asks, "What is the price of product A?"

[2077] 2. The device sends this question to the server.

[2078] 3. The server receives the question and sends it to the generative AI and emotion engine.

[2079] 4. The generative AI generates the answer, "The price of Product A is 10,000 yen."

[2080] 5. The emotion engine recognizes the user's emotions and does not detect negative emotions.

[2081] 6. The server receives this response and determines that there is no problem with the verification.

[2082] 7. The server provides the user with the answer, "The price of product A is 10,000 yen."

[2083] 8. The device displays the answer to the user.

[2084] Example 2: When confirmation is required

[2085] 1. A user asks, "What is the stock status of product B?"

[2086] 2. The device sends this question to the server.

[2087] 3. The server receives the question and sends it to the generative AI and emotion engine.

[2088] 4. The generative AI generates the answer, "The stock of product B is unknown. You may need to check."

[2089] 5. The emotion engine recognizes when a user is anxious.

[2090] 6. The server sets an escalation flag.

[2091] 7. The server sends the inquiry and generated response to the sales representative.

[2092] 8. The sales representative checks the inventory status and replies to the server, "Product B is in stock."

[2093] 9. The server provides the final answer to the user.

[2094] 10. The terminal displays to the user, "Product B is in stock."

[2095] In this way, the present invention is a system that provides users with quick and accurate answers through appropriate collaboration between generative AI, an emotion engine, and sales representatives, and also realizes flexible responses that correspond to the user's emotions.

[2096] Prompt Sentence Examples

[2097] "Please tell me the price of product A."

[2098] "Please let me know the stock status of product B."

[2099] "I would like to know the detailed specifications of product C."

[2100] By using the above, it is possible to respond appropriately and quickly to a wide variety of questions from users.

[2101] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2102] Step 1:

[2103] The user inputs a question such as "What is the price of product A?"

[2104] Specific actions: A user enters a question into an input form in a web browser or mobile application and presses the "Submit" button.

[2105] Input: Text input: "What is the price of product A?"

[2106] Output: The question is sent from the device to the server.

[2107] Step 2:

[2108] The terminal sends this question to the server.

[2109] Specific operation: The terminal sends the user's input to the server as an HTTP request.

[2110] Input: The question entered by the user.

[2111] Output: The question is sent to the server.

[2112] Step 3:

[2113] The server receives the user's query.

[2114] Specific operation: The server parses the HTTP request and extracts the question.

[2115] Input: The question sent from the device.

[2116] Output: The query is ready to be processed by the server.

[2117] Step 4:

[2118] The server sends the question content to the emotion engine and recognizes the emotion.

[2119] Specific operation: The server passes the question content to the emotion engine API and obtains the emotion score.

[2120] Input: The question.

[2121] Output: The emotion score returned by the emotion engine (e.g., happy, anger, sadness, etc.).

[2122] Step 5:

[2123] The server sends the question to the generative AI, which generates an answer.

[2124] How it works: The server calls the generative AI's API and inputs the question. The generative AI retrieves relevant information from its internal database and generates an appropriate answer.

[2125] Input: The question.

[2126] Output: The answer returned by the generative AI (e.g., "Product A costs 10,000 yen").

[2127] Step 6:

[2128] The server verifies the accuracy of the generated answer.

[2129] What happens: The server uses internal logic to check whether the generated answer is of the correct format and content, e.g., whether the price information is in the correct format.

[2130] Input: Answer from a generative AI.

[2131] Output: Answers verified for accuracy.

[2132] Step 7:

[2133] The server adjusts the response content based on the emotion recognized by the emotion engine.

[2134] Specific behavior: The server takes into account the emotion score from the emotion engine and adjusts the style and content of the response. For example, if a negative emotion is detected, it will change the wording to be more polite.

[2135] Input: Sentiment score, generated answer.

[2136] Output: The adjusted answer.

[2137] Step 8:

[2138] The server sets an escalation flag if necessary and sends the inquiry and generated response to a sales representative.

[2139] What it does: If the server is unsure about the generated answer or if the sentiment engine recognizes a negative sentiment, it sets an escalation flag and sends the inquiry and answer to a sales rep.

[2140] Input: A tailored response, and an escalation flag if necessary.

[2141] Output: Enquiry and response sent to sales representative.

[2142] Step 9:

[2143] Sales representatives receive escalated questions and answers and can review and correct them.

[2144] What happens next? Salespeople review details in a dedicated interface (web dashboard or mobile app) and enter revised responses as needed.

[2145] Input: Escalated question and answer.

[2146] Output: The final revised answer.

[2147] Step 10:

[2148] After receiving the final response from the sales representative, the server provides the response to the user.

[2149] Specific operation: The server sends the final answer to the terminal as an HTTP response.

[2150] Input: Final response from sales representative.

[2151] Output: The final answer sent to the user.

[2152] Step 11:

[2153] The terminal displays the final answer to the user.

[2154] Specific operation: The device receives the response from the server and displays the final answer in the user interface.

[2155] Input: The final response from the server.

[2156] Output: The final answer that is displayed to the user.

[2157] The above are the specific processing steps of this system, which allows for quick and accurate answers and flexible responses that reflect the user's emotions.

[2158] (Application example 2)

[2159] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2160] Traditional corporate sales activities and customer support systems face challenges in providing prompt and accurate responses to user inquiries. Furthermore, a lack of response that reflects user emotions poses a risk of a poor user experience. In particular, if users with negative emotions are not properly addressed, customer satisfaction may decline, adversely affecting business growth. To solve these issues, a system that combines generative AI and emotion recognition technology and has an appropriate escalation function is needed.

[2161] 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.

[2162] In this invention, the server includes a means for receiving questions from users, a means for generating answers using a generative AI model, a means for using an emotion engine to recognize user emotions, a means for escalating if a negative emotion is recognized, and a means for providing an answer to the user after final confirmation by a sales representative. This enables quick and accurate responses to user inquiries and flexible responses according to emotions.

[2163] A "server" is a computer system that provides data and services in response to requests from other computers or devices.

[2164] "User" refers to any individual or legal entity that uses the system or services.

[2165] A "Question" is text or audio information provided by a User in the form of an inquiry or request for information.

[2166] A "generative AI model" is an artificial intelligence technology that automatically generates answers based on questions posed by users.

[2167] "Answer" refers to the information or solution provided by a generative AI model or a salesperson.

[2168] "Accuracy" refers to information or answers that are factual and free from errors.

[2169] An "emotion engine" is a technology that recognizes emotions from a user's questions or input, and provides appropriate responses and responses based on those emotions.

[2170] "Negative affect" refers to negative emotional states such as anxiety, doubt, anger, and sadness.

[2171] "Escalation" is the process of passing a difficult problem or situation to a higher level person or expert.

[2172] "Salesperson" is the staff member responsible for communicating with the user and confirming or correcting the final response.

[2173] An "internal database" is a data storage system installed within a company that a generative AI model accesses and retrieves information from.

[2174] This invention is a system for quickly and accurately responding to inquiries in corporate sales activities and customer support systems. In particular, it aims to improve the user experience by identifying the user's emotions and providing answers or escalations based on those emotions. The specific configuration and operation of this system are described in detail below.

[2175] System configuration

[2176] The system consists of the following main components:

[2177] 1. Server

[2178] The server receives questions from users, generates answers using a generative AI model, recognizes user emotions using an emotion engine, and escalates the situation as needed. The server also provides the final answer.

[2179] 2. Generative AI Models

[2180] The generative AI model retrieves information from an internal database and generates appropriate answers based on user questions. The generative AI model is built using natural language processing technology.

[2181] 3. Emotion Engine

[2182] The emotion engine recognizes emotions from user questions and inputs, especially negative emotions (e.g., anxiety or doubt), and adjusts the answers of the generative AI model.

[2183] 4. Escalation function

[2184] If the emotion engine recognizes a negative emotion or if the generated response is unsatisfactory, the server automatically escalates the inquiry and sends the generated response to a sales representative.

[2185] 5. Salesperson Interface

[2186] The sales representative receives the escalated questions and answers, checks and corrects them, and the final corrected answers are provided to the user.

[2187] Hardware and software used

[2188] Smartphone: Used to accept input from users and display results.

[2189] Server: Receives questions, generates answers, recognizes emotions, and handles escalation.

[2190] Generative AI model: Obtains information from an internal database and generates answers.

[2191] Emotion engine: Recognizes user emotions and tailors responses.

[2192] Specific examples

[2193] 1. Receiving user questions

[2194] A user uses a smartphone app to ask, "Can I return product X?" This question is sent to the server.

[2195] 2. Generating answers to questions

[2196] The generative AI model generates the answer, "Returns of product X are generally possible. Click here for details."

[2197] 3. User Emotion Recognition

[2198] The emotion engine recognizes the emotion of doubt from the user's question.

[2199] 4. Escalation

[2200] The server escalates based on the results of the emotion engine and sends the question and AI answer to the sales representative.

[2201] 5. Final confirmation and response

[2202] The sales representative will check the request and provide the final answer to the user via the server: "Returns are possible. If you have any questions, please contact our support center." The answer will be displayed on the smartphone.

[2203] Prompt Sentence Examples

[2204] Prompt for the AI ​​model to perform sentiment analysis:

[2205] Please analyze the following text and classify the emotion: 'Can I return product X?'

[2206] Prompt for generating an answer:

[2207] Generate a customer support response for the following question: 'Can I return product X?' using the company database.

[2208] In this way, the present invention integrates generative AI, an emotion engine, and an escalation function to realize an inquiry response system that can respond flexibly and appropriately according to the user's emotions.

[2209] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2210] Step 1:

[2211] The user inputs and sends a question using a device (smartphone app). An example of a user input is, "Can I return product X?" This input is sent to the server.

[2212] Step 2:

[2213] The server receives the question from the user, which is received in text format and stored internally for further processing.

[2214] Step 3:

[2215] The server sends the received question to the emotion engine, which recognizes the user's emotion from the question. For example, the emotion engine recognizes doubt from the question "Are returns possible?" The processing result is returned to the server.

[2216] Step 4:

[2217] The server sends the question and the emotion data obtained from the emotion engine to the generative AI model. The generative AI model retrieves the corresponding information from its internal database and generates an answer to the question. For example, it generates an answer such as, "Returns of product X are usually possible. Click here for details." This answer is then returned to the server.

[2218] Step 5:

[2219] The server verifies the answer obtained from the generative AI model. During this verification process, data from an internal database is referenced to confirm the accuracy and appropriateness of the answer. If there are no concerns as a result of the verification, the server proceeds to the next step.

[2220] Step 6:

[2221] The server uses data from the emotion engine to determine if the user has a negative emotion, and if a negative emotion is recognized, the server sets an escalation flag.

[2222] Step 7:

[2223] If a negative sentiment is detected, the server automatically escalates the query and sends the generated response to a sales representative.

[2224] Step 8:

[2225] The sales representative checks the received question and the AI-generated answer and modifies it as necessary. For example, if the answer is modified to "Returns are possible. If you have any questions, please contact our support center," the modified answer is returned to the server.

[2226] Step 9:

[2227] The server receives the final confirmed answer from the sales representative and provides it to the user. This answer is displayed on the device (smartphone app) and notified to the user.

[2228] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits 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.

[2229] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with 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.

[2230] 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 robot 414.

[2231] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[2232] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2233] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[2234] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[2235] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[2236] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[2237] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[2238] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2239] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2240] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[2241] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[2242] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[2243] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[2244] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[2245] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[2246] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[2247] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[2248] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[2249] The following is further disclosed regarding the above embodiment.

[2250] (Claim 1)

[2251] a means for receiving questions from users;

[2252] A means for generating an answer to the question using a generative AI;

[2253] means for verifying the accuracy of the generated answers;

[2254] A means for escalating any concerns raised by the verification;

[2255] A means to provide the answer to the user after final confirmation by the sales representative;

[2256] A system including:

[2257] (Claim 2)

[2258] The system of claim 1, wherein the generative AI obtains information from an internal database.

[2259] (Claim 3)

[2260] The system of claim 1, wherein the escalation to the sales representative is performed automatically.

[2261] "Example 1"

[2262] (Claim 1)

[2263] a means for receiving inquiries from users;

[2264] means for generating an answer to the inquiry using a generative AI model;

[2265] means for verifying the accuracy of the generated answers;

[2266] a means for setting an escalation flag if there is any doubt about the accuracy of the answer as a result of the verification;

[2267] means for transmitting the inquiry and the generated response to a sales representative when the escalation flag is set;

[2268] a means for providing the final answer to the user after final confirmation by the sales representative;

[2269] A system including:

[2270] (Claim 2)

[2271] The system of claim 1 , wherein the generative AI model obtains information from an internal database.

[2272] (Claim 3)

[2273] The system of claim 1, wherein the escalation to the sales representative is performed automatically.

[2274] "Application Example 1"

[2275] (Claim 1)

[2276] means for receiving a query from a user;

[2277] A means for generating an answer to the question using a generative AI;

[2278] means for verifying the accuracy of the generated answers;

[2279] A means for escalating any concerns raised by the verification;

[2280] a means for providing an answer to the user after final confirmation by a sales representative;

[2281] a means for displaying the answer on a terminal installed in a physical store;

[2282] A system including:

[2283] (Claim 2)

[2284] The system of claim 1, wherein the generative AI retrieves information from an internal database and generates an appropriate prompt sentence based on the user's question.

[2285] (Claim 3)

[2286] The system of claim 1, wherein escalation to the sales representative is automatically performed and an escalation decision is made after the generative AI generates an answer.

[2287] "Example 2: Combining Emotion Engines"

[2288] (Claim 1)

[2289] a means for receiving questions from users;

[2290] A means for generating an answer to the question using a generative AI;

[2291] means for verifying the accuracy of the generated answers;

[2292] an emotion engine means for recognizing an emotion of a user based on the content of the question;

[2293] a means for adjusting the content of the response generated based on the emotion recognized by the emotion engine;

[2294] means for escalating as necessary as a result of said verification and emotion recognition;

[2295] A means to provide the answer to the user after final confirmation by the sales representative;

[2296] A system including:

[2297] (Claim 2)

[2298] The system of claim 1, wherein the generative AI obtains information from an internal database.

[2299] (Claim 3)

[2300] The system of claim 1, wherein the escalation to the sales representative is performed automatically.

[2301] "Application example 2 when combining emotion engines"

[2302] (Claim 1)

[2303] a means for receiving questions from users;

[2304] A means for generating an answer to the question using a generative AI model;

[2305] means for verifying the accuracy of the generated answers;

[2306] means for using an emotion engine to recognize the emotion of the user;

[2307] means for escalating when a negative emotion is recognized by the emotion engine;

[2308] A means to provide the answer to the user after final confirmation by the sales representative;

[2309] A system including:

[2310] (Claim 2)

[2311] The system of claim 1 , wherein the generative AI model obtains information from an internal database.

[2312] (Claim 3)

[2313] The system of claim 1, wherein the escalation to the sales representative is performed automatically. [Explanation of symbols]

[2314] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for receiving questions from users; A means for generating an answer to the question using a generative AI; means for verifying the accuracy of the generated answers; A means for escalating any concerns raised by the verification; A means to provide the answer to the user after final confirmation by the sales representative; A system including:

2. The system of claim 1 , wherein the generative AI obtains information from an internal database.

3. The system of claim 1 , wherein the escalation to the sales representative is performed automatically.

Images