System

Abstract

A system is provided.SOLUTION: A system including means for receiving candidate data and a scout mail template, extracting individual information from the candidate data, means for inputting the individual information and the scout mail template into a generation AI model to generate a customized scout mail, means for transmitting the generated scout mail to a predetermined candidate, and means for causing a user to confirm the content of the scout mail.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including 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 recent years, there has been a demand for more efficient recruitment processes, but current methods and systems have the problem that creating individual scout emails and conducting interviews takes a lot of time and effort. Performing these processes manually is extremely inefficient, especially when dealing with large amounts of candidate data. Furthermore, during interviews, there are challenges, such as the difficulty of coordinating interviewer schedules and generating appropriate questions in real time. These issues make it difficult to hire excellent talent at the right time. To address this situation, a new system is needed that can effectively automate the generation of scout emails and conduct interviews on behalf of candidates in real time. [Means for solving the problem]

[0005] The present invention is a system including: means for receiving candidate data and a scout email template and extracting individual information from the candidate data; means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email; means for sending the generated scout email to a specific candidate; and means for allowing a user to review the content of the scout email. The present invention also includes a system including means for providing a real-time chat interface and receiving an interview proxy request; means for analyzing the candidate's responses in real time and generating the next question; and means for analyzing the candidate's response data and generating a report after the interview is completed. Furthermore, by including a software module that automatically performs at least part of the generation, sending, and review of the scout email, the overall recruitment process is made more efficient.

[0006] "Candidate Data" means data containing information about a candidate for employment, including, for example, name, contact details, work history, skills, qualifications, and other assessable attributes.

[0007] A "scouting email template" is a draft email created according to a specific format and structure, which can be completed as a personalized scouting email by adding individual candidate information.

[0008] A "generative AI model" refers to an algorithm and its execution environment that uses artificial intelligence technology to make predictions and generate data from input data, and primarily utilizes machine learning and deep learning techniques.

[0009] A "customized scouting email" is an email generated based on a specific template and individual information about a candidate, and includes content specific to that candidate.

[0010] The "real-time chat interface" is a user interface that allows candidates to directly interact with the system, enabling responses and questions to be exchanged on the spot.

[0011] An "interview proxy request" refers to a request from a company or user to the system to conduct an interview on their behalf, and typically includes information such as candidate information, interview date and time, and position details.

[0012] "Real-time analysis" refers to the process of processing received data and information immediately on the spot and obtaining analysis results immediately.

[0013] A "software module" is a portion of a program designed to perform a specific function, operating as part of an overall system but capable of independently performing specific tasks. [Brief explanation of the drawings]

[0014] [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

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] This invention is a system that utilizes an automatic response function using generative AI to automatically generate native-sounding scouting emails and conduct real-time interviews on behalf of candidates. Specific embodiments for implementing this system are described below.

[0036] Automatic generation of scout emails

[0037] 1. Data Acquisition

[0038] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[0039] Server: Receives the data provided by the user and stores it in a database.

[0040] 2. Preparing input data

[0041] Server: Accesses stored candidate data and scouting email templates via API.

[0042] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[0043] 3. Generate scout emails

[0044] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[0045] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[0046] 4. Sending emails

[0047] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify them if necessary.

[0048] User: Check the contents of the scout email and approve its sending.

[0049] Server: Sends approved scouting emails to the designated candidates.

[0050] Real-time interview proxy

[0051] 1. Receiving an interview request

[0052] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[0053] Server: Receives interview requests and stores them in a database.

[0054] 2. Preparing the chat interface

[0055] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[0056] Terminal (Candidate): Access the chat interface you were invited to.

[0057] 3. The interview begins

[0058] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hi [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[0059] 4. Real-time response processing

[0060] Terminal (candidate): Answer the questions.

[0061] Server: Receives and analyzes the candidate's responses using natural language processing techniques to evaluate the content and tone of the applicant's responses.

[0062] Server: Based on the analysis results, the generative AI model generates the next question.

[0063] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[0064] 5. Interview completion and report generation

[0065] Server: End the chat session once the scheduled interview time is up.

[0066] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[0067] 6. Send report

[0068] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0069] The system of the present invention makes it possible to automatically generate scouting emails and conduct interviews in real time, thereby improving the efficiency of the entire recruitment process and enabling companies to recruit excellent personnel quickly and effectively.

[0070] The processing flow will be explained below.

[0071] Automatic generation of scout emails

[0072] Step 1: Data Acquisition

[0073] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[0074] Server: Receives the data provided by the user and stores it in a database.

[0075] Step 2: Prepare input data

[0076] Server: Accesses stored candidate data and scouting email templates via API.

[0077] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[0078] Step 3: Generate a scout email

[0079] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[0080] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[0081] Step 4: Check and send email

[0082] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify it if necessary.

[0083] User: Check the contents of the scout email and approve sending it.

[0084] Server: Sends approved scouting emails to designated candidates.

[0085] Real-time interview proxy

[0086] Step 1: Receiving an interview request

[0087] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[0088] Server: Receives interview requests and stores them in a database.

[0089] Step 2: Preparing the chat interface

[0090] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[0091] Terminal (Candidate): Access the chat interface you were invited to.

[0092] Step 3: The interview begins

[0093] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[0094] Step 4: Real-time response processing

[0095] Terminal (candidate): Answer the questions.

[0096] Server: Receives and analyzes the candidate's responses using natural language processing technology to evaluate the content and tone of the applicant's responses.

[0097] Server: Based on the analysis results, the generative AI model generates the next question.

[0098] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[0099] Step 5: End the interview and generate a report

[0100] Server: Ends the chat session once the scheduled interview time has expired.

[0101] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[0102] Step 6: Submit report

[0103] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0104] Example 1

[0105] 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."

[0106] In the traditional recruitment process, companies face the challenge of efficiently managing communication with candidates and interviews. In particular, the enormous amount of time and effort required to process large amounts of candidate data, send appropriately customized scouting emails, and conduct interviews in real time reduces the efficiency of recruitment.

[0107] 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.

[0108] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a specified candidate, means for allowing a user to confirm the content of the scout email, and means for preprocessing input data and converting it into a format suitable for the generative AI model. This automates the creation and sending of scout emails, enabling efficient and effective approaches to candidates.

[0109] The server also includes a means for providing a real-time chat interface and receiving an interview proxy request, a means for analyzing the candidate's responses in real time and generating a next question, a means for analyzing the candidate's response data and generating a report after the interview is completed, a means for sending a notification to the candidate through the real-time chat interface, and a means for storing the interview request in a database, thereby streamlining the entire interview process and facilitating real-time interview handling and subsequent evaluation.

[0110] "Candidate Data" means data containing information about candidates who are evaluated during the recruitment process, including personal information such as name, contact details, work history, and skills.

[0111] A "scouting email template" is a basic format for a scouting email sent to a candidate, and includes placeholders for dynamically inserting candidate information.

[0112] A "generative AI model" is an artificial intelligence model for text generation and natural language generation, which uses deep learning algorithms to generate output text from input data.

[0113] "Individual Information" is specific information extracted from candidate data, such as name, contact details, work history, and skills.

[0114] "Preprocessing" is the process of converting data into a format suitable for generative AI models, and includes operations such as cleaning and normalizing the data.

[0115] The "real-time chat interface" is a chat system that allows candidates and interviewers to communicate in real time, providing an environment where questions and answers can be exchanged instantly.

[0116] An "interview proxy request" is a request made by a user to the system, and includes information necessary to proxy interview a candidate (such as the candidate's name, interview date and time, and position details).

[0117] "Natural language processing technology" is a technology for understanding, analyzing, and generating human language, and is capable of performing processes such as semantic analysis and sentiment analysis of text.

[0118] "Notification" refers to a communication sent by the system to a candidate or user, which may be in the form of email, SMS, etc.

[0119] The present invention is a system that utilizes a generative AI model to automatically generate scouting emails and conduct interviews in real time. Specific embodiments for implementing this system are described below.

[0120] Automatic generation of scout emails

[0121] In this system, the user first inputs candidate data and a scouting email template into the system. Candidate data includes name, contact information, work history, skills, etc. This data is entered using a web form or a dedicated application.

[0122] The server receives this data and stores it in a database. During this process, data integrity is also checked. Next, the server references the stored candidate data and scouting email templates via API and performs data preprocessing. During preprocessing, the data is cleaned, normalized, and converted into a format suitable for generative AI models.

[0123] The preprocessed data is then fed into a generative AI model, such as a Transformer-based natural language generation model (e.g., GPT-3®). The generative AI model generates a customized scouting email based on the prompt you provide. For example, the generated email might read, "Hello [candidate name], after reviewing your profile, we believe you would be a great fit for our data scientist position. We would love to speak with you."

[0124] The generated scout email is temporarily stored and an interface is provided for the user to review. Through this interface, the user can check the content of the email and modify it if necessary. Finally, once the user approves the sending of the email, the server sends the approved scout email to the specified candidate. Emails are sent using the SMTP protocol or a cloud-based email API (e.g., SendGrid, Amazon SES).

[0125] Real-time interview proxy

[0126] Next, we will explain the real-time interview proxy process. The user sends an interview proxy request to the system. This request includes the candidate's name, interview date and time, and details of the position. The interview proxy request is sent through a dedicated web application or mobile application.

[0127] The server stores the received interview requests in a database and generates a real-time chat interface using real-time communication technologies such as WebRTC and Socket.io, through which candidates are notified of the interview date and time.

[0128] The candidate accesses the designated chat interface and begins the interview. At the start of the interview, the server selects an initial question from a list of basic questions prepared in advance and sends it to the candidate through the chat interface. For example, a question might be sent: "Hello [candidate's name], thank you for taking the time out of your busy schedule. First, could you tell me about your experience so far?"

[0129] The candidate answers the questions, and the answers are sent to the server. The server uses natural language processing technology to analyze the candidate's responses and evaluates their content. Based on the analysis results, a generative AI model generates the next question and sends it to the candidate in real time. This process is repeated until the end of the interview.

[0130] Once the scheduled interview time is over, the server ends the chat session, analyzes the candidate's response data, and generates a report. This report includes an evaluation of the candidate's response content, tone, and skills. Finally, the generated report is sent to the user. For example, a report may be generated stating, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0131] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0132] Automatic generation of scout emails

[0133] Step 1:

[0134] Input: The user enters candidate data (name, contact details, work history, skills, etc.) and a scouting email template into the system.

[0135] How it works: A user enters candidate data using a web form or dedicated application.

[0136] Output: The entered candidate data and scout email template are sent to the server.

[0137] Step 2:

[0138] Input: The server receives candidate data and a scout email template.

[0139] Operation: The server receives this data and saves it in the database. It checks the data integrity when saving the data.

[0140] Output: Saved candidate data and scouting email template.

[0141] Step 3:

[0142] Input: The server references candidate data and scouting email templates stored in the database.

[0143] How it works: The server retrieves this data through an API, cleans and normalizes it, and converts it into a format suitable for generative AI models.

[0144] Output: Pre-processed candidate data and scout email template.

[0145] Step 4:

[0146] Input: Pre-processed candidate data and scout email template.

[0147] How it works: The server feeds data into a generative AI model (e.g., GPT-3) to generate a customized scouting email based on the prompt text.

[0148] Output: The generated customized scout email.

[0149] Step 5:

[0150] Input: The generated customized scout email.

[0151] Behavior: The server temporarily stores the generated scout email and provides an interface for the user to review it.

[0152] Output: The scout email displayed on the user-facing interface.

[0153] Step 6:

[0154] Input: The user checks the contents of the scout email through the interface.

[0155] How it works: The user can review the contents of the scout email in the interface and make any necessary corrections.

[0156] Output: The corrected or approved Scout email.

[0157] Step 7:

[0158] Input: Approved scout email.

[0159] How it works: The server sends approved scouting emails to the selected candidates using the SMTP protocol or a cloud-based email API (e.g., SendGrid, Amazon SES).

[0160] Output: Scout email sent to candidate.

[0161] Real-time interview proxy

[0162] Step 1:

[0163] Input: The user submits an interview request to the system (candidate name, interview date and time, position details).

[0164] How it works: A user submits a request for an interview through a dedicated web or mobile application.

[0165] Output: The interview request sent to the server.

[0166] Step 2:

[0167] Input: The interview request submitted.

[0168] Operation: The server stores the received interview proxy request in a database.

[0169] Output: Interview requests saved in the database.

[0170] Step 3:

[0171] Input: Interview requests saved in the database.

[0172] How it works: The server generates a real-time chat interface and notifies candidates when their interview is scheduled. The chat interface uses technologies such as WebRTC and Socket.io.

[0173] Output: Interview notice sent to candidate.

[0174] Step 4:

[0175] Input: Candidates who have received interview notices.

[0176] How it works: The candidate accesses the designated chat interface and begins the interview.

[0177] Output: Candidates who accessed the chat interface.

[0178] Step 5:

[0179] Input: The candidate accessing the chat interface.

[0180] How it works: The server selects an initial question from a list of basic questions and sends it to the candidate through a chat interface, such as "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience so far?"

[0181] Output: The initial questions sent to the candidate.

[0182] Step 6:

[0183] Input: Candidate's answers to initial questions.

[0184] How it works: The candidate answers questions and sends their answers to the server.

[0185] Output: Candidate's answers sent to the server.

[0186] Step 7:

[0187] Input: Candidate's answer.

[0188] How it works: The server uses natural language processing techniques to analyze the candidate's responses and evaluate their content.

[0189] Output: Parsed candidate response data.

[0190] Step 8:

[0191] Input: Parsed candidate response data.

[0192] How it works: Based on the analysis results, the server uses a generative AI model to generate the next question and send it to the candidate in real time.

[0193] Output: The next question sent to the candidate.

[0194] Step 9:

[0195] Input: Candidate response data at the end of the interview.

[0196] How it works: Once the allotted interview time has elapsed, the server ends the chat session, analyzes the candidate's response data, and generates a report.

[0197] Output: The generated interview report.

[0198] Step 10:

[0199] Input: The generated interview report.

[0200] Behavior: The server sends the generated report to the user. For example, a report might be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0201] Output: Interview report sent to the user.

[0202] (Application example 1)

[0203] 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."

[0204] In the current recruitment process, creating scouting emails and arranging interviews requires a great deal of effort, placing a heavy burden on companies. Furthermore, as recruitment activities increasingly take place in virtual stores and virtual reality environments, traditional methods are no longer effective. Therefore, a new system is needed to automatically generate scouting emails and efficiently conduct real-time interviews in virtual reality environments.

[0205] 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.

[0206] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a predetermined candidate, means for allowing a user to confirm the content of the scout email, means for displaying the scout email in a virtual reality environment, and means for inputting candidate data and confirming the scout email using a virtual reality head-mounted display, thereby enabling automatic generation of scout emails and effective recruitment activities in a virtual reality environment.

[0207] "Candidate Data" means information about an applicant, including their name, contact details, employment history, skills, and experience.

[0208] A "scout email template" is an email template that contains predefined formats and phrases.

[0209] "Individual Information" is specific details extracted from candidate data.

[0210] A "generative AI model" is an artificial intelligence model that performs transformer-based natural language generation.

[0211] A "customized scouting email" is a custom-made scouting email targeted at a specific candidate, generated based on individual information.

[0212] A "virtual reality environment" is an environment that uses computer graphics and related technologies to immerse a user in a virtual scene or situation.

[0213] A "virtual reality head-mounted display" is a headset-type device that allows a user to visually experience a virtual reality environment.

[0214] A "real-time chat interface" is a user interface for exchanging messages instantly over the Internet.

[0215] An "interview proxy request" is a request from a company or user to the system to conduct an interview on their behalf.

[0216] "Response data" refers to data that includes answers and statements provided by candidates during interviews.

[0217] A "report" is a document that analyzes the results of the interview and includes an overall evaluation and recommendations.

[0218] This invention provides a system for automatically generating scouting emails and conducting real-time interviews using a generative AI model. This system is primarily comprised of a server, a terminal, and a user working together.

[0219] Automatic generation of scout emails

[0220] The server receives candidate data and scouting email templates provided by users and stores them in a database. Candidate data includes names, contact details, work history, skills, and experience. The server then preprocesses this data and converts it into a format suitable for input to a generative AI model. The generative AI model used here is a Transformer-based natural language generation model.

[0221] By inputting the preprocessed data into a generative AI model, an individually customized scout email is generated. For example, a scout email may be automatically generated that reads, "Hello [candidate's name], after looking at your profile, I think you would be a perfect fit for our data scientist position. I would be happy to speak with you." The server temporarily stores this generated scout email for the user and provides an interface for reviewing its contents. Once the user has reviewed and revised it and approved it, the server sends the scout email to the specified candidate.

[0222] Furthermore, the system can display the scout email in a virtual reality environment, allowing users to input candidate data and view the generated scout email using a virtual reality head-mounted display.

[0223] Real-time interview proxy

[0224] The server receives a request for an interview from a user and notifies the candidate of a real-time chat interface at the specified date and time. This chat interface is provided in a virtual reality environment. When the candidate accesses the chat interface, the server selects the first question from a pre-prepared question list and sends it to the candidate.

[0225] For example, a question such as "Hello [candidate's name], thank you for taking the time out of your busy schedule. First, could you tell us about your experience so far?" is sent. When the candidate answers the question, the server analyzes the response in real time, generates the next question, and sends it to the candidate. This is how the interview progresses. When the interview is over, the server analyzes the candidate's response data and generates a report including an overall evaluation. The generated report is sent to the user and includes content such as "The interview results for candidate 'Yamada' are as follows. His experience and skill set are evaluated as being very consistent with our expectations."

[0226] With the above configuration, the system of the present invention can automatically generate scouting emails and conduct effective recruitment activities in a virtual reality environment, thereby improving the efficiency of the entire recruitment process and enabling companies to quickly and effectively recruit talented people.

[0227] Prompt Sentence Examples

[0228] As a concrete example, the following prompt sentence can be considered:

[0229] "Hi [candidate's name], I'm interested in your skill set... I'd love to speak with you."

[0230] In this way, generative AI models and virtual reality environments can be used to recruit more effectively and quickly than traditional methods.

[0231] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0232] Step 1:

[0233] The server receives the candidate data and scouting email template provided by the user and stores it in a database. The input here is the candidate data (name, contact details, work history, skills, etc.) and the scouting email template, and the output is the stored data, which will be used in later processing steps.

[0234] Step 2:

[0235] The server references the stored candidate data and scout email templates and performs preprocessing. Specifically, it cleans and normalizes the data and converts it into a format suitable for the generative AI model. The input is the candidate data and scout email templates, and the output is the preprocessed data. This data is used as input to the generative AI model.

[0236] Step 3:

[0237] The server inputs the preprocessed candidate data and the scout email template into the generative AI model. The generative AI model generates a customized scout email based on the input data. Here, the input is the preprocessed data, and the output is the generated scout email. This scout email contains content that is tailored to each candidate.

[0238] Step 4:

[0239] The server temporarily stores the generated scout email and provides a confirmation interface to the user. The user can check the contents of the scout email through this interface and modify it as necessary. The input is the generated scout email, and the output is the scout email that the user has checked and modified.

[0240] Step 5:

[0241] After the user confirms, modifies, and approves the scout email, the server sends the approved scout email to the specified candidate. The input is the scout email approved by the user, and the output is the sent scout email. This sends the scout email to the candidate.

[0242] Step 6:

[0243] The user uses a virtual reality head-mounted display to input candidate data and view the generated scout email in the virtual reality environment. The input is the scout email displayed in the virtual reality environment, and the output is the scout email viewed by the user.

[0244] Step 7:

[0245] The server receives a request for a job interview from a user and notifies the candidate via a real-time chat interface at the specified date and time. The input is the job interview request and the output is the notification to the candidate. This chat interface is provided within a virtual reality environment.

[0246] Step 8:

[0247] The server manages the interview process through a real-time chat interface. First, it selects initial questions from a pre-prepared question list and sends them to the candidate. The input is the question list, and the output is the questions sent to the candidate.

[0248] Step 9:

[0249] The terminal (candidate) answers questions through a real-time chat interface, and the responses are sent to the server. The input is the candidate's answer, and the output is the response data sent to the server.

[0250] Step 10:

[0251] The server analyzes the candidate's response data and generates the next question. Natural language processing technology is used for the analysis. The input is the candidate's response data, and the output is the generated next question. This question generation is done in real time and is repeated until the end of the interview.

[0252] Step 11:

[0253] Once the interview is over, the server analyzes the candidate's response data and generates a report that includes an overall evaluation of the interview. The input is the response data, and the output is the generated report. This report is sent to the user and includes, for example, "The interview results for candidate 'Yamada' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0254] 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.

[0255] The present invention is a system that recognizes and responds to user emotions in the automatic generation of native-sounding scouting emails and real-time interview proxy by combining an automatic response function using generative AI with an emotion engine. Specific embodiments for implementing this system are described below.

[0256] Automatic generation of scout emails

[0257] 1. Data Acquisition

[0258] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[0259] Server: Receives the data provided by the user and stores it in a database.

[0260] 2. Preparing input data

[0261] Server: Accesses stored candidate data and scouting email templates via API.

[0262] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[0263] 3. Generate scout emails

[0264] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[0265] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[0266] 4. Content adjustment based on sentiment analysis

[0267] Server: Using the emotion engine, the server estimates the emotional state of the potential candidate based on the generated scouting email.

[0268] Server: Optimize the content of the scouting email based on the candidate's estimated emotional state. For example, if the candidate is estimated to be nervous, add more relaxing language.

[0269] 5. Check and send emails

[0270] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify it if necessary.

[0271] User: Check the contents of the scout email and approve sending it.

[0272] Server: Sends approved scouting emails to designated candidates.

[0273] Real-time interview proxy

[0274] 1. Receiving an interview request

[0275] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[0276] Server: Receives interview requests and stores them in a database.

[0277] 2. Preparing the chat interface

[0278] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[0279] Terminal (Candidate): Access the chat interface you were invited to.

[0280] 3. The interview begins

[0281] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[0282] 4. Real-time response processing

[0283] Terminal (candidate): Answer the questions.

[0284] Server: Receives and analyzes the candidate's responses using natural language processing technology to evaluate the content and tone of the applicant's responses.

[0285] Server: Based on the analysis results, the emotion engine estimates the candidate's emotional state. For example, if it estimates that the candidate is nervous, it adjusts the next question to make them feel more relaxed.

[0286] Server: The generative AI model generates the next question.

[0287] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[0288] 5. Interview completion and report generation

[0289] Server: Ends the chat session once the scheduled interview time has expired.

[0290] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[0291] 6. Send report

[0292] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0293] The system of the present invention enables automatic generation of scouting emails and real-time interviews, and by combining this with sentiment analysis, it provides a higher level of personalization and improves the candidate experience, allowing companies to recruit excellent talent quickly and effectively.

[0294] The processing flow will be explained below.

[0295] Automatic generation of scout emails

[0296] Step 1: Data Acquisition

[0297] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[0298] Server: Receives the data provided by the user and stores it in a database.

[0299] Step 2: Prepare input data

[0300] Server: Accesses stored candidate data and scouting email templates via API.

[0301] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, specifically by cleaning and normalizing the data.

[0302] Step 3: Generate a scout email

[0303] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[0304] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and think you'd be a great fit for our position. We'd love to speak with you."

[0305] Step 4: Content adjustment through sentiment analysis

[0306] Server: Uses an emotion engine on the generated scouting email to estimate the emotional state of the potential candidate.

[0307] Server: Optimize the content of the scouting email based on the estimated emotional state. For example, if it is estimated that the candidate is easily nervous, add expressions that will help them relax.

[0308] Step 5: Check and send email

[0309] Server: Temporarily stores the generated scout email for the user to review. Provides an interface that allows the user to review the contents of the email and modify it if necessary.

[0310] User: Check the contents of the scout email and approve sending it.

[0311] Server: Sends approved scouting emails to designated candidates.

[0312] Real-time interview proxy

[0313] Step 1: Receiving an interview request

[0314] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[0315] Server: Receives interview requests and stores them in a database.

[0316] Step 2: Preparing the chat interface

[0317] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[0318] Terminal (Candidate): Access the chat interface you were invited to.

[0319] Step 3: The interview begins

[0320] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, a question might be something like, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[0321] Step 4: Real-time response processing

[0322] Terminal (candidate): Answer the questions.

[0323] Server: Receives and analyzes candidate responses using natural language processing techniques to evaluate the content and tone of the applicant's responses.

[0324] Server: Based on the analysis results, the emotion engine estimates the candidate's emotional state. For example, if the candidate is estimated to be nervous, the next question will be adjusted to make them feel more relaxed.

[0325] Server: The generative AI model generates the next question.

[0326] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[0327] Step 5: End the interview and generate a report

[0328] Server: Ends the chat session once the scheduled interview time has expired.

[0329] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[0330] Step 6: Submit report

[0331] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0332] Example 2

[0333] 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."

[0334] The traditional process of creating scouting emails and conducting interviews required a great deal of time and effort, and it was difficult to properly consider candidates' emotions and reactions. Furthermore, manually creating scouting emails and conducting interviews was inefficient and difficult to respond to individually, preventing the rapid and effective recruitment of excellent talent. Furthermore, the inability to respond to candidates' emotions can lead to a poor candidate experience, putting companies at a disadvantage in the competition for talent.

[0335] 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.

[0336] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for analyzing the generated scout email with an emotion engine and adjusting the content based on the candidate's emotional state, means for sending the adjusted scout email to a specified candidate, and means for allowing a user to confirm the content of the scout email. This enables automatic generation of scout emails and adjustment of the content based on emotions.

[0337] The system also includes a means for providing a real-time chat interface and receiving an interview proxy request, a means for analyzing the candidate's responses in real time and estimating the candidate's emotional state using an emotion engine, a means for using a generative AI model to generate the next question based on the candidate's emotional state, and a means for analyzing the candidate's response data and generating a report after the interview is completed, thereby enabling real-time interview proxy and appropriate responses according to the candidate's emotions.

[0338] "Candidate Data" means data that includes information such as a candidate's name, contact details, work history, and skills.

[0339] A "scout email template" is a template containing a standard email format and structure used to generate scout emails.

[0340] "Generative AI models" are artificial intelligence models that automatically generate text based on input data, including transformer-based natural language generation models.

[0341] An "emotion engine" is an engine that uses text analysis technology to estimate the emotional state of a target and adjusts responses and content based on that.

[0342] A "real-time chat interface" is an interface that enables synchronous communication, and is a system that allows real-time interaction in chat format.

[0343] An "interview proxy request" is a request to the system to proxy an interview, and includes information such as the candidate's name, interview date and time, and position details.

[0344] A "report" is a document generated after the interview is completed that summarizes information including the candidate's response data, sentiment analysis results, and overall evaluation.

[0345] A "software module" is a portion of program code designed to perform a specific function, such as at least partially automatically generating, sentiment analyzing, content adjusting, sending, and reviewing scouting emails.

[0346] The present invention is a system that combines a generative AI model and an emotion engine to automatically generate scouting emails and conduct real-time interviews. Specific embodiments of the present invention are described in detail below.

[0347] Automatic generation of scout emails

[0348] 1. Data Acquisition

[0349] User: The user enters candidate data (such as name, contact details, work history, and skills) and a scouting email template into the system's input form. For example, the user enters "Taro Tanaka" as the candidate's name.

[0350] Server: The server receives the candidate data and email template provided by the user, converts them into JSON format, and stores them in a NoSQL database.

[0351] 2. Preparing input data

[0352] Server: Retrieves stored candidate data and scouting email templates from the database via API. Executes SQL queries to extract the required data.

[0353] Server: Preprocesses the acquired data and converts it into a format suitable for generative AI models. For example, it cleans and normalizes the data, outputs a warning log for missing data, and sets default values.

[0354] 3. Generate scout emails

[0355] Server: Input the preprocessed candidate data and templates into a Transformer-based generative AI model (e.g., GPT-3). The prompt sentence is "Generate a scouting email based on the skill set of candidate 'Taro Tanaka'."

[0356] Server: Receives a customized scouting email generated by the AI ​​model, such as "Hello, Taro Tanaka, after looking at your profile, I think you would be a great fit for our data scientist position. I would be happy to speak with you."

[0357] 4. Content adjustment based on sentiment analysis

[0358] Server: The generated scouting email is sent to the emotion engine, which estimates the candidate's emotional state when they receive it. The content of the email is analyzed using NLP technology.

[0359] Server: Adjust the email's language based on the candidate's estimated emotional state. For example, if the candidate is estimated to be nervous, add a relaxing message.

[0360] 5. Check and send emails

[0361] Server: The generated scout email is temporarily saved and displayed on a dashboard that can be viewed by the user.

[0362] User: The user checks the email on the dashboard, makes any necessary corrections, and once the final check is complete, clicks the send button to approve it.

[0363] Server: Automatically sends approved scouting emails to candidates using the SMTP protocol.

[0364] Real-time interview proxy

[0365] 1. Receiving an interview request

[0366] User: The user enters a request for an interview on the system's dedicated form. The request includes the candidate's name, the scheduled interview date and time, and the desired interview content. For example, the user might enter "Taro Tanaka, interview for a data scientist position, October 1, 2023, 10:00 AM" as the request content.

[0367] Server: The server receives the request data from the user, converts it into JSON format, and saves it in the database.

[0368] 2. Preparing the chat interface

[0369] Server: Generates a real-time chat interface at the specified date and time and sends a notification email to the candidate, including a link to access the interview.

[0370] Device (Candidate): The candidate receives a notification email and clicks on the link to access the chat interface.

[0371] 3. The interview begins

[0372] Server: At the start of the interview, select an initial question from a basic list and send it to the candidate via the chat interface. For example, "Hello, Taro Tanaka. First, please tell me about your experience so far."

[0373] 4. Real-time response processing

[0374] Terminal (candidate): Candidate answers questions through a chat interface.

[0375] Server: Analyzes the received answers in real time and uses an emotion engine to estimate the candidate's emotional state, for example, detecting stress or tension from the tone and content of the answers.

[0376] Server: Generate the next question based on the emotional state. Using a generative AI model, enter the prompt "Generate the next question for the interview with candidate 'Taro Tanaka'."

[0377] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the interview is over.

[0378] 5. Interview completion and report generation

[0379] Server: Ends the chat session once the scheduled interview time has expired.

[0380] Server: Analyzes the interview data and generates a report including an overall evaluation of the interview based on the candidate's response data and sentiment analysis results. For example, it generates a report such as, "The interview results for candidate 'Taro Tanaka' are as follows. His experience and skill set are excellent and meet our expectations."

[0381] 6. Send report

[0382] Server: Automatically sends generated reports to users, either displaying them in a dedicated dashboard or sending them via email.

[0383] In this way, the present invention significantly improves the efficiency of the process of automatically generating scouting emails and conducting real-time interviews, and makes it possible to respond appropriately to the emotions of candidates, thereby enabling companies to recruit excellent talent quickly and effectively.

[0384] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0385] Automatic generation of scout emails

[0386] Step 1: Data Acquisition

[0387] User: Enter candidate data (name, contact details, work history, skills, etc.) and a scouting email template into the input form.

[0388] Input: Candidate data, email template

[0389] Output: User input data

[0390] Specific actions: Enter the required candidate information and template into the system's input form and click the submit button.

[0391] Server: Receives user input data, converts it to JSON format, and saves it in the database.

[0392] Input: User-entered data

[0393] Output: Candidate data stored in database, email template

[0394] Specific operation: Analyzes the received data, converts it into an appropriate format, and stores it in a NoSQL database.

[0395] Step 2: Prepare input data

[0396] Server: Uses API to retrieve candidate data and scout email templates from the database.

[0397] Input: Database query

[0398] Output: Retrieved candidate data, email template

[0399] Specific behavior: Executes SQL queries and extracts stored data.

[0400] Server: Preprocesses the acquired data, complements missing data, and normalizes the data.

[0401] Input: Captured candidate data, email template

[0402] Output: Preprocessed data

[0403] Specific operations: Data is cleaned, normalized, missing data is filled, and the data is converted into a format suitable for generative AI models.

[0404] Step 3: Generate a scout email

[0405] Server: The preprocessed data is input into a generative AI model (e.g., GPT-3) to generate a scouting email.

[0406] Input: Preprocessed data, prompt "Generate a scouting email based on the skill set of candidate 'Taro Tanaka'"

[0407] Output: Generated scout email

[0408] Specific operation: Send data including a prompt sentence to the generative AI model and receive the generated email.

[0409] Step 4: Content adjustment through sentiment analysis

[0410] Server: The generated scouting email is sent to the emotion engine, and the candidate's emotional state upon receiving it is estimated.

[0411] Input: Generated scout email

[0412] Output: Estimated emotional state

[0413] Specific operation: The emotion engine analyzes the content of the email and obtains emotion estimation results.

[0414] Server: Adjusts the wording of the email based on the estimated emotional state.

[0415] Input: Generated scout email, emotion estimation result

[0416] Output: Adjusted scout email

[0417] Specific operation: The content of the email is adjusted according to the emotion estimation results, and relaxing expressions are added.

[0418] Step 5: Check and send email

[0419] Server: Temporarily stores generated and adjusted scout emails and displays them on a dashboard that users can view.

[0420] Input: Adjusted Scout Mail

[0421] Output: Scout email displayed on the dashboard

[0422] Specific behavior: Display the email content in the dashboard interface.

[0423] User: Check the email on the dashboard, make any corrections, and once the final check is complete, click the send button.

[0424] Input: Scout email displayed on the dashboard

[0425] Output: User approved scout email

[0426] Specific actions: Check the email content, make any necessary corrections, and click the send button.

[0427] Server: Sends approved scouting emails to candidates.

[0428] Input: User-approved scout email

[0429] Output: Scout email sent

[0430] What it does: Automatically sends emails using the SMTP protocol.

[0431] Real-time interview proxy

[0432] Step 1: Receiving an interview request

[0433] User: Enters a request for interview on a dedicated form, along with the desired interview date and time and candidate information.

[0434] Input: Interview request data (candidate name, interview date and time, position details, etc.)

[0435] Output: User input data

[0436] Specific actions: Enter the required information in the request form and click the submit button.

[0437] Server: Receives request data from the user and stores it in a database.

[0438] Input: User-entered data

[0439] Output: Request data stored in the database

[0440] Specific operation: Convert the received data into an appropriate format and save it in the database.

[0441] Step 2: Preparing the chat interface

[0442] Server: Generates a real-time chat interface at the specified date and time and sends a notification email to the candidate.

[0443] Input: Request data (interview date and time, candidate contact information)

[0444] Output: Generated chat interface, notification email

[0445] What it does: Generates an email containing a chat interface link and sends it to the candidate.

[0446] Terminal (Candidate): The candidate receives a notification email and accesses the chat interface.

[0447] Input: Notification email, access link

[0448] Output: Chat interface accessed

[0449] Specific action: Click on the link in the email to enter the chat interface.

[0450] Step 3: The interview begins

[0451] Server: At the start of the interview, select initial questions from a basic list and send them to the candidate via a chat interface.

[0452] Input: A list of pre-prepared questions

[0453] Output: Initial question submitted

[0454] What it does: Select an appropriate question from the list of questions and send it to the chat interface. For example, "Hello, Taro Tanaka. First, please tell me about your experience."

[0455] Step 4: Real-time response processing

[0456] Terminal (candidate): Candidate answers questions through a chat interface.

[0457] Input: Initial question

[0458] Output: Candidate's answers

[0459] Specific actions: Enter a response into the chat interface screen and send it.

[0460] Server: Analyzes received answers in real time and uses an emotion engine to estimate the candidate's emotional state.

[0461] Input: Candidate's answer

[0462] Output: Emotional state estimation result

[0463] What it does: It uses natural language processing techniques to analyze responses and infer emotional states, for example detecting stress or tension from the tone and content of responses.

[0464] Server: Generates the next question based on the emotional state.

[0465] Input: Emotional state estimation result, previous answer

[0466] Output: Next question

[0467] Specific operation: The generative AI model is given the prompt "Generate the next question for the interview with candidate 'Taro Tanaka'" and the next question is created.

[0468] Server: Generates the next question and sends it to the candidate in real time.

[0469] Input: Next question

[0470] Output: Question submitted

[0471] What it does: It sends the generated questions to the chat interface and displays them to the candidate. This process is repeated until the interview is over.

[0472] Step 5: End the interview and generate a report

[0473] Server: Ends the chat session once the scheduled interview time has expired.

[0474] Input: Interview End Notice

[0475] Output: Chat sessions completed

[0476] Specific behavior: When the interview end time is reached, a notification will be sent automatically to end the chat session.

[0477] Server: Analyzes the overall interview data and generates a report including an overall evaluation.

[0478] Input: Interview response data, emotion analysis results

[0479] Output: Generated report

[0480] Specific behavior: Create an interview evaluation report based on the candidate's response data and sentiment analysis results. For example, "The interview results for candidate 'Taro Tanaka' are as follows. His experience and skill set match our expectations."

[0481] Step 6: Submit report

[0482] Server: Automatically sends generated reports to users.

[0483] Input: Generated report

[0484] Output: Report sent

[0485] What it does: View the report in a dedicated dashboard or send it via email.

[0486] (Application example 2)

[0487] 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."

[0488] Traditional recruitment and customer service processes have struggled to provide personalized experiences because they struggle to respond appropriately to the emotions of candidates and customers. There is also a need for a way to improve customer satisfaction by understanding customer interests and emotions in real time and generating responses based on them. However, traditional systems lack an efficient means to meet these needs.

[0489] 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.

[0490] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a specified candidate, means for allowing a user to confirm the content of the scout email, means for analyzing the emotional state of a customer in real time via a smart device, means for generating an appropriate response to the customer using the generative AI model based on the analysis results, and means for displaying the generated response on the smart device. This makes it possible to efficiently generate and send personalized responses based on the emotions of candidates and customers.

[0491] "Candidate Data" refers to specific information about people involved in the recruitment process, such as their names, contact details, work history, and skills.

[0492] A "scouting email template" refers to a template that includes the basic structure and sample text of a scouting email to be sent to candidates.

[0493] "Individual Information" refers to specific pieces of information extracted from candidate data.

[0494] A "generative AI model" is an artificial intelligence model that generates natural language based on data, such as a Transformer-based model like GPT-3.

[0495] "Customized scouting email" refers to a scouting email that is tailored to a specific candidate based on the candidate's individual information.

[0496] "Smart devices" refers to devices that can connect to the Internet, such as smartphones, smart glasses, and head-mounted displays.

[0497] "Emotional state" refers to a state that indicates a client's or candidate's psychological reactions and feelings.

[0498] "Analysis results" refers to the results of data processing generated based on emotional state and other input data.

[0499] An "appropriate response" refers to the optimal reply or message that is generated based on the situation and the other person's emotional state.

[0500] "Real-time chat interface" refers to a user interface that enables real-time text and voice communication over the Internet.

[0501] This invention is a system that combines a smart device, a generative AI model, and a sentiment analysis engine to generate personalized responses for candidate and customer interactions. Specific embodiments for implementing this system are described below.

[0502] Hardware and software used

[0503] Hardware: Smartphones, smart glasses, head-mounted displays, cameras

[0504] Software: OpenCV, transformers library (GPT-3 model), EmotionEngine (emotion analysis engine)

[0505] System Program Processing

[0506] The server first receives candidate data and a scout email template and extracts individual information from the candidate data. It then inputs the individual information and the scout email template into a generative AI model to generate a customized scout email. It then sends the generated scout email to the designated candidate and has the user review its content. During this process, the server analyzes the customer's emotional state in real time via the smart device, and generates an appropriate response for the customer using the generative AI model based on the emotion analysis results, which is then displayed on the smart device.

[0507] Specific examples

[0508] For example, a store staff member wearing smart glasses captures a customer's facial expression with a camera and analyzes the image data using Emotion Engine. If Emotion Engine detects that the customer is interested as a result of the analysis, the server sends a prompt to the generative AI model, such as "The customer is interested. Please explain this product to us." The generated response, such as "This product is the latest model, with especially enhanced durability. Please try it out," is displayed in real time on the smart glasses. This allows the store staff member to immediately provide the customer with an appropriate explanation.

[0509] Prompt Sentence Examples

[0510] "The customer is smiling. Can you tell me how to approach them in a relaxed manner?"

[0511] "My customer is looking at my product with interest. What's the best way to approach them and sell them this product?"

[0512] "Customer wants more information. Can you give me a detailed description of this product?"

[0513] The system enables efficient generation and delivery of personalized responses based on candidate and customer sentiment, resulting in a better candidate and customer experience, helping companies hire top talent quickly and effectively and increasing customer satisfaction.

[0514] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0515] Step 1:

[0516] The user enters candidate data and a scouting email template into the system. The entered candidate data includes details such as name, contact information, work history, and skills. This data is sent to the server.

[0517] Step 2:

[0518] The server stores the received candidate data and scouting email templates, makes the stored data available via API, and then performs pre-processing to extract specific information, including cleaning and normalizing the data.

[0519] Step 3:

[0520] The server inputs the preprocessed candidate data and the scouting email template into a generative AI model to generate a customized scouting email. For example, a Transformer-based natural language generation model (GPT-3) is used to generate an email such as, "Hello [candidate name], we've looked at your profile and believe you'd be a great fit for our position. We'd love to speak with you."

[0521] Step 4:

[0522] The server provides a means for the user to check the generated scout email before sending it to a specific candidate. The user checks the contents of the scout email, corrects it if necessary, and then approves its sending.

[0523] Step 5:

[0524] After receiving approval from the user, the server sends a scouting email to the candidate. This sending process is automatic.

[0525] Step 6:

[0526] When a customer visits the store, the device (a user wearing smart glasses or a head-mounted display) captures their facial expressions and movements with a camera, and the captured video data is sent to the server in real time.

[0527] Step 7:

[0528] The server uses the Emotion Engine to analyze the video data sent and determine the customer's emotional state, resulting in an analysis result such as "This customer is showing interest."

[0529] Step 8:

[0530] Based on the results of the emotional state analysis, the server inputs a prompt to the generative AI model, such as "The customer is interested. Please explain this product to me."

[0531] Step 9:

[0532] The generative AI model generates an appropriate response based on the input prompt. The generated response is sent back to the server. For example, a response such as, "This product is the latest model and has been especially strengthened for durability. Please try it out."

[0533] Step 10:

[0534] The server then displays the generated response on smart glasses or a head-mounted display, allowing the user (store staff) to provide appropriate explanations to the customer in real time.

[0535] The input and output at each step is through data processing and data calculations, allowing personalized responses based on candidate or customer sentiment to be automatically generated and sent.

[0536] 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.

[0537] 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.

[0538] 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.

[0539] [Second embodiment]

[0540] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0541] 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.

[0542] 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).

[0543] 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.

[0544] 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.

[0545] 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).

[0546] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for 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.

[0547] 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.

[0548] 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.

[0549] 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.

[0550] 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.

[0551] 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."

[0552] This invention is a system that utilizes an automatic response function using generative AI to automatically generate native-sounding scouting emails and conduct real-time interviews on behalf of candidates. Specific embodiments for implementing this system are described below.

[0553] Automatic generation of scout emails

[0554] 1. Data Acquisition

[0555] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[0556] Server: Receives the data provided by the user and stores it in a database.

[0557] 2. Preparing input data

[0558] Server: Accesses stored candidate data and scouting email templates via API.

[0559] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[0560] 3. Generate scout emails

[0561] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[0562] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[0563] 4. Sending emails

[0564] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify them if necessary.

[0565] User: Check the contents of the scout email and approve its sending.

[0566] Server: Sends approved scouting emails to the designated candidates.

[0567] Real-time interview proxy

[0568] 1. Receiving an interview request

[0569] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[0570] Server: Receives interview requests and stores them in a database.

[0571] 2. Preparing the chat interface

[0572] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[0573] Terminal (Candidate): Access the chat interface you were invited to.

[0574] 3. The interview begins

[0575] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hi [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[0576] 4. Real-time response processing

[0577] Terminal (candidate): Answer the questions.

[0578] Server: Receives and analyzes the candidate's responses using natural language processing techniques to evaluate the content and tone of the applicant's responses.

[0579] Server: Based on the analysis results, the generative AI model generates the next question.

[0580] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[0581] 5. Interview completion and report generation

[0582] Server: End the chat session once the scheduled interview time is up.

[0583] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[0584] 6. Send report

[0585] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0586] The system of the present invention makes it possible to automatically generate scouting emails and conduct interviews in real time, thereby improving the efficiency of the entire recruitment process and enabling companies to recruit excellent personnel quickly and effectively.

[0587] The processing flow will be explained below.

[0588] Automatic generation of scout emails

[0589] Step 1: Data Acquisition

[0590] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[0591] Server: Receives the data provided by the user and stores it in a database.

[0592] Step 2: Prepare input data

[0593] Server: Accesses stored candidate data and scouting email templates via API.

[0594] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[0595] Step 3: Generate a scout email

[0596] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[0597] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[0598] Step 4: Check and send email

[0599] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify it if necessary.

[0600] User: Check the contents of the scout email and approve sending it.

[0601] Server: Sends approved scouting emails to designated candidates.

[0602] Real-time interview proxy

[0603] Step 1: Receiving an interview request

[0604] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[0605] Server: Receives interview requests and stores them in a database.

[0606] Step 2: Preparing the chat interface

[0607] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[0608] Terminal (Candidate): Access the chat interface you were invited to.

[0609] Step 3: The interview begins

[0610] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[0611] Step 4: Real-time response processing

[0612] Terminal (candidate): Answer the questions.

[0613] Server: Receives and analyzes the candidate's responses using natural language processing technology to evaluate the content and tone of the applicant's responses.

[0614] Server: Based on the analysis results, the generative AI model generates the next question.

[0615] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[0616] Step 5: End the interview and generate a report

[0617] Server: Ends the chat session once the scheduled interview time has expired.

[0618] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[0619] Step 6: Submit report

[0620] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0621] Example 1

[0622] 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."

[0623] In the traditional recruitment process, companies face the challenge of efficiently managing communication with candidates and interviews. In particular, the enormous amount of time and effort required to process large amounts of candidate data, send appropriately customized scouting emails, and conduct interviews in real time reduces the efficiency of recruitment.

[0624] 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.

[0625] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a specified candidate, means for allowing a user to confirm the content of the scout email, and means for preprocessing input data and converting it into a format suitable for the generative AI model. This automates the creation and sending of scout emails, enabling efficient and effective approaches to candidates.

[0626] The server also includes a means for providing a real-time chat interface and receiving an interview proxy request, a means for analyzing the candidate's responses in real time and generating a next question, a means for analyzing the candidate's response data and generating a report after the interview is completed, a means for sending a notification to the candidate through the real-time chat interface, and a means for storing the interview request in a database, thereby streamlining the entire interview process and facilitating real-time interview handling and subsequent evaluation.

[0627] "Candidate Data" means data containing information about candidates who are evaluated during the recruitment process, including personal information such as name, contact details, work history, and skills.

[0628] A "scouting email template" is a basic format for a scouting email sent to a candidate, and includes placeholders for dynamically inserting candidate information.

[0629] A "generative AI model" is an artificial intelligence model for text generation and natural language generation, which uses deep learning algorithms to generate output text from input data.

[0630] "Individual Information" is specific information extracted from candidate data, such as name, contact details, work history, and skills.

[0631] "Preprocessing" is the process of converting data into a format suitable for generative AI models, and includes operations such as cleaning and normalizing the data.

[0632] The "real-time chat interface" is a chat system that allows candidates and interviewers to communicate in real time, providing an environment where questions and answers can be exchanged instantly.

[0633] An "interview proxy request" is a request made by a user to the system, and includes information necessary to proxy interview a candidate (such as the candidate's name, interview date and time, and position details).

[0634] "Natural language processing technology" is a technology for understanding, analyzing, and generating human language, and is capable of performing processes such as semantic analysis and sentiment analysis of text.

[0635] "Notification" refers to a communication sent by the system to a candidate or user, which may be in the form of email, SMS, etc.

[0636] The present invention is a system that utilizes a generative AI model to automatically generate scouting emails and conduct interviews in real time. Specific embodiments for implementing this system are described below.

[0637] Automatic generation of scout emails

[0638] In this system, the user first inputs candidate data and a scouting email template into the system. Candidate data includes name, contact information, work history, skills, etc. This data is entered using a web form or a dedicated application.

[0639] The server receives this data and stores it in a database. During this process, data integrity is also checked. Next, the server references the stored candidate data and scouting email templates via API and performs data preprocessing. During preprocessing, the data is cleaned, normalized, and converted into a format suitable for generative AI models.

[0640] The preprocessed data is then fed into a generative AI model, such as a Transformer-based natural language generation model (e.g., GPT-3). The generative AI model generates a customized scouting email based on the prompt you provide. For example, it might say, "Hello [candidate name], after reviewing your profile, we believe you would be a great fit for our data scientist position. We would love to speak with you."

[0641] The generated scout email is temporarily stored and an interface is provided for the user to review. Through this interface, the user can check the content of the email and modify it if necessary. Finally, once the user approves the sending of the email, the server sends the approved scout email to the specified candidate. Emails are sent using the SMTP protocol or a cloud-based email API (e.g., SendGrid, Amazon SES).

[0642] Real-time interview proxy

[0643] Next, we will explain the real-time interview proxy process. The user sends an interview proxy request to the system. This request includes the candidate's name, interview date and time, and details of the position. The interview proxy request is sent through a dedicated web application or mobile application.

[0644] The server stores the received interview requests in a database and generates a real-time chat interface using real-time communication technologies such as WebRTC and Socket.io, through which candidates are notified of the interview date and time.

[0645] The candidate accesses the designated chat interface and begins the interview. At the start of the interview, the server selects an initial question from a list of basic questions prepared in advance and sends it to the candidate through the chat interface. For example, a question might be sent: "Hello [candidate's name], thank you for taking the time out of your busy schedule. First, could you tell me about your experience so far?"

[0646] The candidate answers the questions, and the answers are sent to the server. The server uses natural language processing technology to analyze the candidate's responses and evaluates their content. Based on the analysis results, a generative AI model generates the next question and sends it to the candidate in real time. This process is repeated until the end of the interview.

[0647] Once the scheduled interview time is over, the server ends the chat session, analyzes the candidate's response data, and generates a report. This report includes an evaluation of the candidate's response content, tone, and skills. Finally, the generated report is sent to the user. For example, a report may be generated stating, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0648] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0649] Automatic generation of scout emails

[0650] Step 1:

[0651] Input: The user enters candidate data (name, contact details, work history, skills, etc.) and a scouting email template into the system.

[0652] How it works: A user enters candidate data using a web form or dedicated application.

[0653] Output: The entered candidate data and scout email template are sent to the server.

[0654] Step 2:

[0655] Input: The server receives candidate data and a scout email template.

[0656] Operation: The server receives this data and saves it in the database. It checks the data integrity when saving the data.

[0657] Output: Saved candidate data and scouting email template.

[0658] Step 3:

[0659] Input: The server references candidate data and scouting email templates stored in the database.

[0660] How it works: The server retrieves this data through an API, cleans and normalizes it, and converts it into a format suitable for generative AI models.

[0661] Output: Pre-processed candidate data and scout email template.

[0662] Step 4:

[0663] Input: Pre-processed candidate data and scout email template.

[0664] How it works: The server feeds data into a generative AI model (e.g., GPT-3) to generate a customized scouting email based on the prompt text.

[0665] Output: The generated customized scout email.

[0666] Step 5:

[0667] Input: The generated customized scout email.

[0668] Behavior: The server temporarily stores the generated scout email and provides an interface for the user to review it.

[0669] Output: The scout email displayed on the user-facing interface.

[0670] Step 6:

[0671] Input: The user checks the contents of the scout email through the interface.

[0672] How it works: The user can review the contents of the scout email in the interface and make any necessary corrections.

[0673] Output: The corrected or approved Scout email.

[0674] Step 7:

[0675] Input: Approved scout email.

[0676] How it works: The server sends approved scouting emails to the selected candidates using the SMTP protocol or a cloud-based email API (e.g., SendGrid, Amazon SES).

[0677] Output: Scout email sent to candidate.

[0678] Real-time interview proxy

[0679] Step 1:

[0680] Input: The user submits an interview request to the system (candidate name, interview date and time, position details).

[0681] How it works: A user submits a request for an interview through a dedicated web or mobile application.

[0682] Output: The interview request sent to the server.

[0683] Step 2:

[0684] Input: The interview request submitted.

[0685] Operation: The server stores the received interview proxy request in a database.

[0686] Output: Interview requests saved in the database.

[0687] Step 3:

[0688] Input: Interview requests saved in the database.

[0689] How it works: The server generates a real-time chat interface and notifies candidates when their interview is scheduled. The chat interface uses technologies such as WebRTC and Socket.io.

[0690] Output: Interview notice sent to candidate.

[0691] Step 4:

[0692] Input: Candidates who have received interview notices.

[0693] How it works: The candidate accesses the designated chat interface and begins the interview.

[0694] Output: Candidates who accessed the chat interface.

[0695] Step 5:

[0696] Input: The candidate accessing the chat interface.

[0697] How it works: The server selects an initial question from a list of basic questions and sends it to the candidate through a chat interface, such as "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience so far?"

[0698] Output: The initial questions sent to the candidate.

[0699] Step 6:

[0700] Input: Candidate's answers to initial questions.

[0701] How it works: The candidate answers questions and sends their answers to the server.

[0702] Output: Candidate's answers sent to the server.

[0703] Step 7:

[0704] Input: Candidate's answer.

[0705] How it works: The server uses natural language processing techniques to analyze the candidate's responses and evaluate their content.

[0706] Output: Parsed candidate response data.

[0707] Step 8:

[0708] Input: Parsed candidate response data.

[0709] How it works: Based on the analysis results, the server uses a generative AI model to generate the next question and send it to the candidate in real time.

[0710] Output: The next question sent to the candidate.

[0711] Step 9:

[0712] Input: Candidate response data at the end of the interview.

[0713] How it works: Once the allotted interview time has elapsed, the server ends the chat session, analyzes the candidate's response data, and generates a report.

[0714] Output: The generated interview report.

[0715] Step 10:

[0716] Input: The generated interview report.

[0717] Behavior: The server sends the generated report to the user. For example, a report might be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0718] Output: Interview report sent to the user.

[0719] (Application example 1)

[0720] 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."

[0721] In the current recruitment process, creating scouting emails and arranging interviews requires a great deal of effort, placing a heavy burden on companies. Furthermore, as recruitment activities increasingly take place in virtual stores and virtual reality environments, traditional methods are no longer effective. Therefore, a new system is needed to automatically generate scouting emails and efficiently conduct real-time interviews in virtual reality environments.

[0722] 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.

[0723] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a predetermined candidate, means for allowing a user to confirm the content of the scout email, means for displaying the scout email in a virtual reality environment, and means for inputting candidate data and confirming the scout email using a virtual reality head-mounted display, thereby enabling automatic generation of scout emails and effective recruitment activities in a virtual reality environment.

[0724] "Candidate Data" means information about an applicant, including their name, contact details, employment history, skills, and experience.

[0725] A "scout email template" is an email template that contains predefined formats and phrases.

[0726] "Individual Information" is specific details extracted from candidate data.

[0727] A "generative AI model" is an artificial intelligence model that performs transformer-based natural language generation.

[0728] A "customized scouting email" is a custom-made scouting email targeted at a specific candidate, generated based on individual information.

[0729] A "virtual reality environment" is an environment that uses computer graphics and related technologies to immerse a user in a virtual scene or situation.

[0730] A "virtual reality head-mounted display" is a headset-type device that allows a user to visually experience a virtual reality environment.

[0731] A "real-time chat interface" is a user interface for exchanging messages instantly over the Internet.

[0732] An "interview proxy request" is a request from a company or user to the system to conduct an interview on their behalf.

[0733] "Response data" refers to data that includes answers and statements provided by candidates during interviews.

[0734] A "report" is a document that analyzes the results of the interview and includes an overall evaluation and recommendations.

[0735] This invention provides a system for automatically generating scouting emails and conducting real-time interviews using a generative AI model. This system is primarily comprised of a server, a terminal, and a user working together.

[0736] Automatic generation of scout emails

[0737] The server receives candidate data and scouting email templates provided by users and stores them in a database. Candidate data includes names, contact details, work history, skills, and experience. The server then preprocesses this data and converts it into a format suitable for input to a generative AI model. The generative AI model used here is a Transformer-based natural language generation model.

[0738] By inputting the preprocessed data into a generative AI model, an individually customized scout email is generated. For example, a scout email may be automatically generated that reads, "Hello [candidate's name], after looking at your profile, I think you would be a perfect fit for our data scientist position. I would be happy to speak with you." The server temporarily stores this generated scout email for the user and provides an interface for reviewing its contents. Once the user has reviewed and revised it and approved it, the server sends the scout email to the specified candidate.

[0739] Furthermore, the system can display the scout email in a virtual reality environment, allowing users to input candidate data and view the generated scout email using a virtual reality head-mounted display.

[0740] Real-time interview proxy

[0741] The server receives a request for an interview from a user and notifies the candidate of a real-time chat interface at the specified date and time. This chat interface is provided in a virtual reality environment. When the candidate accesses the chat interface, the server selects the first question from a pre-prepared question list and sends it to the candidate.

[0742] For example, a question such as "Hello [candidate's name], thank you for taking the time out of your busy schedule. First, could you tell us about your experience so far?" is sent. When the candidate answers the question, the server analyzes the response in real time, generates the next question, and sends it to the candidate. This is how the interview progresses. When the interview is over, the server analyzes the candidate's response data and generates a report including an overall evaluation. The generated report is sent to the user and includes content such as "The interview results for candidate 'Yamada' are as follows. His experience and skill set are evaluated as being very consistent with our expectations."

[0743] With the above configuration, the system of the present invention can automatically generate scouting emails and conduct effective recruitment activities in a virtual reality environment, thereby improving the efficiency of the entire recruitment process and enabling companies to quickly and effectively recruit talented people.

[0744] Prompt Sentence Examples

[0745] As a concrete example, the following prompt sentence can be considered:

[0746] "Hi [candidate's name], I'm interested in your skill set... I'd love to speak with you."

[0747] In this way, generative AI models and virtual reality environments can be used to recruit more effectively and quickly than traditional methods.

[0748] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0749] Step 1:

[0750] The server receives the candidate data and scouting email template provided by the user and stores it in a database. The input here is the candidate data (name, contact details, work history, skills, etc.) and the scouting email template, and the output is the stored data, which will be used in later processing steps.

[0751] Step 2:

[0752] The server references the stored candidate data and scout email templates and performs preprocessing. Specifically, it cleans and normalizes the data and converts it into a format suitable for the generative AI model. The input is the candidate data and scout email templates, and the output is the preprocessed data. This data is used as input to the generative AI model.

[0753] Step 3:

[0754] The server inputs the preprocessed candidate data and the scout email template into the generative AI model. The generative AI model generates a customized scout email based on the input data. Here, the input is the preprocessed data, and the output is the generated scout email. This scout email contains content that is tailored to each candidate.

[0755] Step 4:

[0756] The server temporarily stores the generated scout email and provides a confirmation interface to the user. The user can check the contents of the scout email through this interface and modify it as necessary. The input is the generated scout email, and the output is the scout email that the user has checked and modified.

[0757] Step 5:

[0758] After the user confirms, modifies, and approves the scout email, the server sends the approved scout email to the specified candidate. The input is the scout email approved by the user, and the output is the sent scout email. This sends the scout email to the candidate.

[0759] Step 6:

[0760] The user uses a virtual reality head-mounted display to input candidate data and view the generated scout email in the virtual reality environment. The input is the scout email displayed in the virtual reality environment, and the output is the scout email viewed by the user.

[0761] Step 7:

[0762] The server receives a request for a job interview from a user and notifies the candidate via a real-time chat interface at the specified date and time. The input is the job interview request and the output is the notification to the candidate. This chat interface is provided within a virtual reality environment.

[0763] Step 8:

[0764] The server manages the interview process through a real-time chat interface. First, it selects initial questions from a pre-prepared question list and sends them to the candidate. The input is the question list, and the output is the questions sent to the candidate.

[0765] Step 9:

[0766] The terminal (candidate) answers questions through a real-time chat interface, and the responses are sent to the server. The input is the candidate's answer, and the output is the response data sent to the server.

[0767] Step 10:

[0768] The server analyzes the candidate's response data and generates the next question. Natural language processing technology is used for the analysis. The input is the candidate's response data, and the output is the generated next question. This question generation is done in real time and is repeated until the end of the interview.

[0769] Step 11:

[0770] Once the interview is over, the server analyzes the candidate's response data and generates a report that includes an overall evaluation of the interview. The input is the response data, and the output is the generated report. This report is sent to the user and includes, for example, "The interview results for candidate 'Yamada' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0771] 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.

[0772] The present invention is a system that recognizes and responds to user emotions in the automatic generation of native-sounding scouting emails and real-time interview proxy by combining an automatic response function using generative AI with an emotion engine. Specific embodiments for implementing this system are described below.

[0773] Automatic generation of scout emails

[0774] 1. Data Acquisition

[0775] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[0776] Server: Receives the data provided by the user and stores it in a database.

[0777] 2. Preparing input data

[0778] Server: Accesses stored candidate data and scouting email templates via API.

[0779] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[0780] 3. Generate scout emails

[0781] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[0782] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[0783] 4. Content adjustment based on sentiment analysis

[0784] Server: Using the emotion engine, the server estimates the emotional state of the potential candidate based on the generated scouting email.

[0785] Server: Optimize the content of the scouting email based on the candidate's estimated emotional state. For example, if the candidate is estimated to be nervous, add more relaxing language.

[0786] 5. Check and send emails

[0787] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify it if necessary.

[0788] User: Check the contents of the scout email and approve sending it.

[0789] Server: Sends approved scouting emails to designated candidates.

[0790] Real-time interview proxy

[0791] 1. Receiving an interview request

[0792] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[0793] Server: Receives interview requests and stores them in a database.

[0794] 2. Preparing the chat interface

[0795] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[0796] Terminal (Candidate): Access the chat interface you were invited to.

[0797] 3. The interview begins

[0798] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[0799] 4. Real-time response processing

[0800] Terminal (candidate): Answer the questions.

[0801] Server: Receives and analyzes the candidate's responses using natural language processing technology to evaluate the content and tone of the applicant's responses.

[0802] Server: Based on the analysis results, the emotion engine estimates the candidate's emotional state. For example, if it estimates that the candidate is nervous, it adjusts the next question to make them feel more relaxed.

[0803] Server: The generative AI model generates the next question.

[0804] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[0805] 5. Interview completion and report generation

[0806] Server: Ends the chat session once the scheduled interview time has expired.

[0807] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[0808] 6. Send report

[0809] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0810] The system of the present invention enables automatic generation of scouting emails and real-time interviews, and by combining this with sentiment analysis, it provides a higher level of personalization and improves the candidate experience, allowing companies to recruit excellent talent quickly and effectively.

[0811] The processing flow will be explained below.

[0812] Automatic generation of scout emails

[0813] Step 1: Data Acquisition

[0814] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[0815] Server: Receives the data provided by the user and stores it in a database.

[0816] Step 2: Prepare input data

[0817] Server: Accesses stored candidate data and scouting email templates via API.

[0818] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, specifically by cleaning and normalizing the data.

[0819] Step 3: Generate a scout email

[0820] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[0821] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and think you'd be a great fit for our position. We'd love to speak with you."

[0822] Step 4: Content adjustment through sentiment analysis

[0823] Server: Uses an emotion engine on the generated scouting email to estimate the emotional state of the potential candidate.

[0824] Server: Optimize the content of the scouting email based on the estimated emotional state. For example, if it is estimated that the candidate is easily nervous, add expressions that will help them relax.

[0825] Step 5: Check and send email

[0826] Server: Temporarily stores the generated scout email for the user to review. Provides an interface that allows the user to review the contents of the email and modify it if necessary.

[0827] User: Check the contents of the scout email and approve sending it.

[0828] Server: Sends approved scouting emails to designated candidates.

[0829] Real-time interview proxy

[0830] Step 1: Receiving an interview request

[0831] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[0832] Server: Receives interview requests and stores them in a database.

[0833] Step 2: Preparing the chat interface

[0834] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[0835] Terminal (Candidate): Access the chat interface you were invited to.

[0836] Step 3: The interview begins

[0837] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, a question might be something like, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[0838] Step 4: Real-time response processing

[0839] Terminal (candidate): Answer the questions.

[0840] Server: Receives and analyzes candidate responses using natural language processing techniques to evaluate the content and tone of the applicant's responses.

[0841] Server: Based on the analysis results, the emotion engine estimates the candidate's emotional state. For example, if the candidate is estimated to be nervous, the next question will be adjusted to make them feel more relaxed.

[0842] Server: The generative AI model generates the next question.

[0843] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[0844] Step 5: End the interview and generate a report

[0845] Server: Ends the chat session once the scheduled interview time has expired.

[0846] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[0847] Step 6: Submit report

[0848] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[0849] Example 2

[0850] 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."

[0851] The traditional process of creating scouting emails and conducting interviews required a great deal of time and effort, and it was difficult to properly consider candidates' emotions and reactions. Furthermore, manually creating scouting emails and conducting interviews was inefficient and difficult to respond to individually, preventing the rapid and effective recruitment of excellent talent. Furthermore, the inability to respond to candidates' emotions can lead to a poor candidate experience, putting companies at a disadvantage in the competition for talent.

[0852] 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.

[0853] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for analyzing the generated scout email with an emotion engine and adjusting the content based on the candidate's emotional state, means for sending the adjusted scout email to a specified candidate, and means for allowing a user to confirm the content of the scout email. This enables automatic generation of scout emails and adjustment of the content based on emotions.

[0854] The system also includes a means for providing a real-time chat interface and receiving an interview proxy request, a means for analyzing the candidate's responses in real time and estimating the candidate's emotional state using an emotion engine, a means for using a generative AI model to generate the next question based on the candidate's emotional state, and a means for analyzing the candidate's response data and generating a report after the interview is completed, thereby enabling real-time interview proxy and appropriate responses according to the candidate's emotions.

[0855] "Candidate Data" means data that includes information such as a candidate's name, contact details, work history, and skills.

[0856] A "scout email template" is a template containing a standard email format and structure used to generate scout emails.

[0857] "Generative AI models" are artificial intelligence models that automatically generate text based on input data, including transformer-based natural language generation models.

[0858] An "emotion engine" is an engine that uses text analysis technology to estimate the emotional state of a target and adjusts responses and content based on that.

[0859] A "real-time chat interface" is an interface that enables synchronous communication, and is a system that allows real-time interaction in chat format.

[0860] An "interview proxy request" is a request to the system to proxy an interview, and includes information such as the candidate's name, interview date and time, and position details.

[0861] A "report" is a document generated after the interview is completed that summarizes information including the candidate's response data, sentiment analysis results, and overall evaluation.

[0862] A "software module" is a portion of program code designed to perform a specific function, such as at least partially automatically generating, sentiment analyzing, content adjusting, sending, and reviewing scouting emails.

[0863] The present invention is a system that combines a generative AI model and an emotion engine to automatically generate scouting emails and conduct real-time interviews. Specific embodiments of the present invention are described in detail below.

[0864] Automatic generation of scout emails

[0865] 1. Data Acquisition

[0866] User: The user enters candidate data (such as name, contact details, work history, and skills) and a scouting email template into the system's input form. For example, the user enters "Taro Tanaka" as the candidate's name.

[0867] Server: The server receives the candidate data and email template provided by the user, converts them into JSON format, and stores them in a NoSQL database.

[0868] 2. Preparing input data

[0869] Server: Retrieves stored candidate data and scouting email templates from the database via API. Executes SQL queries to extract the required data.

[0870] Server: Preprocesses the acquired data and converts it into a format suitable for generative AI models. For example, it cleans and normalizes the data, outputs a warning log for missing data, and sets default values.

[0871] 3. Generate scout emails

[0872] Server: Input the preprocessed candidate data and templates into a Transformer-based generative AI model (e.g., GPT-3). The prompt sentence is "Generate a scouting email based on the skill set of candidate 'Taro Tanaka'."

[0873] Server: Receives a customized scouting email generated by the AI ​​model, such as "Hello, Taro Tanaka, after looking at your profile, I think you would be a great fit for our data scientist position. I would be happy to speak with you."

[0874] 4. Content adjustment based on sentiment analysis

[0875] Server: The generated scouting email is sent to the emotion engine, which estimates the candidate's emotional state when they receive it. The content of the email is analyzed using NLP technology.

[0876] Server: Adjust the email's language based on the candidate's estimated emotional state. For example, if the candidate is estimated to be nervous, add a relaxing message.

[0877] 5. Check and send emails

[0878] Server: The generated scout email is temporarily saved and displayed on a dashboard that can be viewed by the user.

[0879] User: The user checks the email on the dashboard, makes any necessary corrections, and once the final check is complete, clicks the send button to approve it.

[0880] Server: Automatically sends approved scouting emails to candidates using the SMTP protocol.

[0881] Real-time interview proxy

[0882] 1. Receiving an interview request

[0883] User: The user enters a request for an interview on the system's dedicated form. The request includes the candidate's name, the scheduled interview date and time, and the desired interview content. For example, the user might enter "Taro Tanaka, interview for a data scientist position, October 1, 2023, 10:00 AM" as the request content.

[0884] Server: The server receives the request data from the user, converts it into JSON format, and saves it in the database.

[0885] 2. Preparing the chat interface

[0886] Server: Generates a real-time chat interface at the specified date and time and sends a notification email to the candidate, including a link to access the interview.

[0887] Device (Candidate): The candidate receives a notification email and clicks on the link to access the chat interface.

[0888] 3. The interview begins

[0889] Server: At the start of the interview, select an initial question from a basic list and send it to the candidate via the chat interface. For example, "Hello, Taro Tanaka. First, please tell me about your experience so far."

[0890] 4. Real-time response processing

[0891] Terminal (candidate): Candidate answers questions through a chat interface.

[0892] Server: Analyzes the received answers in real time and uses an emotion engine to estimate the candidate's emotional state, for example, detecting stress or tension from the tone and content of the answers.

[0893] Server: Generate the next question based on the emotional state. Using a generative AI model, enter the prompt "Generate the next question for the interview with candidate 'Taro Tanaka'."

[0894] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the interview is over.

[0895] 5. Interview completion and report generation

[0896] Server: Ends the chat session once the scheduled interview time has expired.

[0897] Server: Analyzes the interview data and generates a report including an overall evaluation of the interview based on the candidate's response data and sentiment analysis results. For example, it generates a report such as, "The interview results for candidate 'Taro Tanaka' are as follows. His experience and skill set are excellent and meet our expectations."

[0898] 6. Send report

[0899] Server: Automatically sends generated reports to users, either displaying them in a dedicated dashboard or sending them via email.

[0900] In this way, the present invention significantly improves the efficiency of the process of automatically generating scouting emails and conducting real-time interviews, and makes it possible to respond appropriately to the emotions of candidates, thereby enabling companies to recruit excellent talent quickly and effectively.

[0901] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0902] Automatic generation of scout emails

[0903] Step 1: Data Acquisition

[0904] User: Enter candidate data (name, contact details, work history, skills, etc.) and a scouting email template into the input form.

[0905] Input: Candidate data, email template

[0906] Output: User input data

[0907] Specific actions: Enter the required candidate information and template into the system's input form and click the submit button.

[0908] Server: Receives user input data, converts it to JSON format, and saves it in the database.

[0909] Input: User-entered data

[0910] Output: Candidate data stored in database, email template

[0911] Specific operation: Analyzes the received data, converts it into an appropriate format, and stores it in a NoSQL database.

[0912] Step 2: Prepare input data

[0913] Server: Uses API to retrieve candidate data and scout email templates from the database.

[0914] Input: Database query

[0915] Output: Retrieved candidate data, email template

[0916] Specific behavior: Executes SQL queries and extracts stored data.

[0917] Server: Preprocesses the acquired data, complements missing data, and normalizes the data.

[0918] Input: Captured candidate data, email template

[0919] Output: Preprocessed data

[0920] Specific operations: Data is cleaned, normalized, missing data is filled, and the data is converted into a format suitable for generative AI models.

[0921] Step 3: Generate a scout email

[0922] Server: The preprocessed data is input into a generative AI model (e.g., GPT-3) to generate a scouting email.

[0923] Input: Preprocessed data, prompt "Generate a scouting email based on the skill set of candidate 'Taro Tanaka'"

[0924] Output: Generated scout email

[0925] Specific operation: Send data including a prompt sentence to the generative AI model and receive the generated email.

[0926] Step 4: Content adjustment through sentiment analysis

[0927] Server: The generated scouting email is sent to the emotion engine, and the candidate's emotional state upon receiving it is estimated.

[0928] Input: Generated scout email

[0929] Output: Estimated emotional state

[0930] Specific operation: The emotion engine analyzes the content of the email and obtains emotion estimation results.

[0931] Server: Adjusts the wording of the email based on the estimated emotional state.

[0932] Input: Generated scout email, emotion estimation result

[0933] Output: Adjusted scout email

[0934] Specific operation: The content of the email is adjusted according to the emotion estimation results, and relaxing expressions are added.

[0935] Step 5: Check and send email

[0936] Server: Temporarily stores generated and adjusted scout emails and displays them on a dashboard that users can view.

[0937] Input: Adjusted Scout Mail

[0938] Output: Scout email displayed on the dashboard

[0939] Specific behavior: Display the email content in the dashboard interface.

[0940] User: Check the email on the dashboard, make any corrections, and once the final check is complete, click the send button.

[0941] Input: Scout email displayed on the dashboard

[0942] Output: User approved scout email

[0943] Specific actions: Check the email content, make any necessary corrections, and click the send button.

[0944] Server: Sends approved scouting emails to candidates.

[0945] Input: User-approved scout email

[0946] Output: Scout email sent

[0947] What it does: Automatically sends emails using the SMTP protocol.

[0948] Real-time interview proxy

[0949] Step 1: Receiving an interview request

[0950] User: Enters a request for interview on a dedicated form, along with the desired interview date and time and candidate information.

[0951] Input: Interview request data (candidate name, interview date and time, position details, etc.)

[0952] Output: User input data

[0953] Specific actions: Enter the required information in the request form and click the submit button.

[0954] Server: Receives request data from the user and stores it in a database.

[0955] Input: User-entered data

[0956] Output: Request data stored in the database

[0957] Specific operation: Convert the received data into an appropriate format and save it in the database.

[0958] Step 2: Preparing the chat interface

[0959] Server: Generates a real-time chat interface at the specified date and time and sends a notification email to the candidate.

[0960] Input: Request data (interview date and time, candidate contact information)

[0961] Output: Generated chat interface, notification email

[0962] What it does: Generates an email containing a chat interface link and sends it to the candidate.

[0963] Terminal (Candidate): The candidate receives a notification email and accesses the chat interface.

[0964] Input: Notification email, access link

[0965] Output: Chat interface accessed

[0966] Specific action: Click on the link in the email to enter the chat interface.

[0967] Step 3: The interview begins

[0968] Server: At the start of the interview, select initial questions from a basic list and send them to the candidate via a chat interface.

[0969] Input: A list of pre-prepared questions

[0970] Output: Initial question submitted

[0971] What it does: Select an appropriate question from the list of questions and send it to the chat interface. For example, "Hello, Taro Tanaka. First, please tell me about your experience."

[0972] Step 4: Real-time response processing

[0973] Terminal (candidate): Candidate answers questions through a chat interface.

[0974] Input: Initial question

[0975] Output: Candidate's answers

[0976] Specific actions: Enter a response into the chat interface screen and send it.

[0977] Server: Analyzes received answers in real time and uses an emotion engine to estimate the candidate's emotional state.

[0978] Input: Candidate's answer

[0979] Output: Emotional state estimation result

[0980] What it does: It uses natural language processing techniques to analyze responses and infer emotional states, for example detecting stress or tension from the tone and content of responses.

[0981] Server: Generates the next question based on the emotional state.

[0982] Input: Emotional state estimation result, previous answer

[0983] Output: Next question

[0984] Specific operation: The generative AI model is given the prompt "Generate the next question for the interview with candidate 'Taro Tanaka'" and the next question is created.

[0985] Server: Generates the next question and sends it to the candidate in real time.

[0986] Input: Next question

[0987] Output: Question submitted

[0988] What it does: It sends the generated questions to the chat interface and displays them to the candidate. This process is repeated until the interview is over.

[0989] Step 5: End the interview and generate a report

[0990] Server: Ends the chat session once the scheduled interview time has expired.

[0991] Input: Interview End Notice

[0992] Output: Chat sessions completed

[0993] Specific behavior: When the interview end time is reached, a notification will be sent automatically to end the chat session.

[0994] Server: Analyzes the overall interview data and generates a report including an overall evaluation.

[0995] Input: Interview response data, emotion analysis results

[0996] Output: Generated report

[0997] Specific behavior: Create an interview evaluation report based on the candidate's response data and sentiment analysis results. For example, "The interview results for candidate 'Taro Tanaka' are as follows. His experience and skill set match our expectations."

[0998] Step 6: Submit report

[0999] Server: Automatically sends generated reports to users.

[1000] Input: Generated report

[1001] Output: Report sent

[1002] What it does: View the report in a dedicated dashboard or send it via email.

[1003] (Application example 2)

[1004] 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."

[1005] Traditional recruitment and customer service processes have struggled to provide personalized experiences because they struggle to respond appropriately to the emotions of candidates and customers. There is also a need for a way to improve customer satisfaction by understanding customer interests and emotions in real time and generating responses based on them. However, traditional systems lack an efficient means to meet these needs.

[1006] 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.

[1007] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a specified candidate, means for allowing a user to confirm the content of the scout email, means for analyzing the emotional state of a customer in real time via a smart device, means for generating an appropriate response to the customer using the generative AI model based on the analysis results, and means for displaying the generated response on the smart device. This makes it possible to efficiently generate and send personalized responses based on the emotions of candidates and customers.

[1008] "Candidate Data" refers to specific information about people involved in the recruitment process, such as their names, contact details, work history, and skills.

[1009] A "scouting email template" refers to a template that includes the basic structure and sample text of a scouting email to be sent to candidates.

[1010] "Individual Information" refers to specific pieces of information extracted from candidate data.

[1011] A "generative AI model" is an artificial intelligence model that generates natural language based on data, such as a Transformer-based model like GPT-3.

[1012] "Customized scouting email" refers to a scouting email that is tailored to a specific candidate based on the candidate's individual information.

[1013] "Smart devices" refers to devices that can connect to the Internet, such as smartphones, smart glasses, and head-mounted displays.

[1014] "Emotional state" refers to a state that indicates a client's or candidate's psychological reactions and feelings.

[1015] "Analysis results" refers to the results of data processing generated based on emotional state and other input data.

[1016] An "appropriate response" refers to the optimal reply or message that is generated based on the situation and the other person's emotional state.

[1017] "Real-time chat interface" refers to a user interface that enables real-time text and voice communication over the Internet.

[1018] This invention is a system that combines a smart device, a generative AI model, and a sentiment analysis engine to generate personalized responses for candidate and customer interactions. Specific embodiments for implementing this system are described below.

[1019] Hardware and software used

[1020] Hardware: Smartphones, smart glasses, head-mounted displays, cameras

[1021] Software: OpenCV, transformers library (GPT-3 model), EmotionEngine (emotion analysis engine)

[1022] System Program Processing

[1023] The server first receives candidate data and a scout email template and extracts individual information from the candidate data. It then inputs the individual information and the scout email template into a generative AI model to generate a customized scout email. It then sends the generated scout email to the designated candidate and has the user review its content. During this process, the server analyzes the customer's emotional state in real time via the smart device, and generates an appropriate response for the customer using the generative AI model based on the emotion analysis results, which is then displayed on the smart device.

[1024] Specific examples

[1025] For example, a store staff member wearing smart glasses captures a customer's facial expression with a camera and analyzes the image data using Emotion Engine. If Emotion Engine detects that the customer is interested as a result of the analysis, the server sends a prompt to the generative AI model, such as "The customer is interested. Please explain this product to us." The generated response, such as "This product is the latest model, with especially enhanced durability. Please try it out," is displayed in real time on the smart glasses. This allows the store staff member to immediately provide the customer with an appropriate explanation.

[1026] Prompt Sentence Examples

[1027] "The customer is smiling. Can you tell me how to approach them in a relaxed manner?"

[1028] "My customer is looking at my product with interest. What's the best way to approach them and sell them this product?"

[1029] "Customer wants more information. Can you give me a detailed description of this product?"

[1030] The system enables efficient generation and delivery of personalized responses based on candidate and customer sentiment, resulting in a better candidate and customer experience, helping companies hire top talent quickly and effectively and increasing customer satisfaction.

[1031] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1032] Step 1:

[1033] The user enters candidate data and a scouting email template into the system. The entered candidate data includes details such as name, contact information, work history, and skills. This data is sent to the server.

[1034] Step 2:

[1035] The server stores the received candidate data and scouting email templates, makes the stored data available via API, and then performs pre-processing to extract specific information, including cleaning and normalizing the data.

[1036] Step 3:

[1037] The server inputs the preprocessed candidate data and the scouting email template into a generative AI model to generate a customized scouting email. For example, a Transformer-based natural language generation model (GPT-3) is used to generate an email such as, "Hello [candidate name], we've looked at your profile and believe you'd be a great fit for our position. We'd love to speak with you."

[1038] Step 4:

[1039] The server provides a means for the user to check the generated scout email before sending it to a specific candidate. The user checks the contents of the scout email, corrects it if necessary, and then approves its sending.

[1040] Step 5:

[1041] After receiving approval from the user, the server sends a scouting email to the candidate. This sending process is automatic.

[1042] Step 6:

[1043] When a customer visits the store, the device (a user wearing smart glasses or a head-mounted display) captures their facial expressions and movements with a camera, and the captured video data is sent to the server in real time.

[1044] Step 7:

[1045] The server uses the Emotion Engine to analyze the video data sent and determine the customer's emotional state, resulting in an analysis result such as "This customer is showing interest."

[1046] Step 8:

[1047] Based on the results of the emotional state analysis, the server inputs a prompt to the generative AI model, such as "The customer is interested. Please explain this product to me."

[1048] Step 9:

[1049] The generative AI model generates an appropriate response based on the input prompt. The generated response is sent back to the server. For example, a response such as, "This product is the latest model and has been especially strengthened for durability. Please try it out."

[1050] Step 10:

[1051] The server then displays the generated response on smart glasses or a head-mounted display, allowing the user (store staff) to provide appropriate explanations to the customer in real time.

[1052] The input and output at each step is through data processing and data calculations, allowing personalized responses based on candidate or customer sentiment to be automatically generated and sent.

[1053] 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.

[1054] 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.

[1055] 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.

[1056] [Third embodiment]

[1057] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1058] 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.

[1059] 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).

[1060] 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.

[1061] 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.

[1062] 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).

[1063] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for 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.

[1064] 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.

[1065] 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.

[1066] 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.

[1067] 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.

[1068] 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."

[1069] This invention is a system that utilizes an automatic response function using generative AI to automatically generate native-sounding scouting emails and conduct real-time interviews on behalf of candidates. Specific embodiments for implementing this system are described below.

[1070] Automatic generation of scout emails

[1071] 1. Data Acquisition

[1072] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[1073] Server: Receives the data provided by the user and stores it in a database.

[1074] 2. Preparing input data

[1075] Server: Accesses stored candidate data and scouting email templates via API.

[1076] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[1077] 3. Generate scout emails

[1078] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[1079] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[1080] 4. Sending emails

[1081] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify them if necessary.

[1082] User: Check the contents of the scout email and approve its sending.

[1083] Server: Sends approved scouting emails to the designated candidates.

[1084] Real-time interview proxy

[1085] 1. Receiving an interview request

[1086] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[1087] Server: Receives interview requests and stores them in a database.

[1088] 2. Preparing the chat interface

[1089] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[1090] Terminal (Candidate): Access the chat interface you were invited to.

[1091] 3. The interview begins

[1092] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hi [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[1093] 4. Real-time response processing

[1094] Terminal (candidate): Answer the questions.

[1095] Server: Receives and analyzes the candidate's responses using natural language processing techniques to evaluate the content and tone of the applicant's responses.

[1096] Server: Based on the analysis results, the generative AI model generates the next question.

[1097] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[1098] 5. Interview completion and report generation

[1099] Server: End the chat session once the scheduled interview time is up.

[1100] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[1101] 6. Send report

[1102] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1103] The system of the present invention makes it possible to automatically generate scouting emails and conduct interviews in real time, thereby improving the efficiency of the entire recruitment process and enabling companies to recruit excellent personnel quickly and effectively.

[1104] The processing flow will be explained below.

[1105] Automatic generation of scout emails

[1106] Step 1: Data Acquisition

[1107] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[1108] Server: Receives the data provided by the user and stores it in a database.

[1109] Step 2: Prepare input data

[1110] Server: Accesses stored candidate data and scouting email templates via API.

[1111] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[1112] Step 3: Generate a scout email

[1113] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[1114] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[1115] Step 4: Check and send email

[1116] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify it if necessary.

[1117] User: Check the contents of the scout email and approve sending it.

[1118] Server: Sends approved scouting emails to designated candidates.

[1119] Real-time interview proxy

[1120] Step 1: Receiving an interview request

[1121] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[1122] Server: Receives interview requests and stores them in a database.

[1123] Step 2: Preparing the chat interface

[1124] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[1125] Terminal (Candidate): Access the chat interface you were invited to.

[1126] Step 3: The interview begins

[1127] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[1128] Step 4: Real-time response processing

[1129] Terminal (candidate): Answer the questions.

[1130] Server: Receives and analyzes the candidate's responses using natural language processing technology to evaluate the content and tone of the applicant's responses.

[1131] Server: Based on the analysis results, the generative AI model generates the next question.

[1132] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[1133] Step 5: End the interview and generate a report

[1134] Server: Ends the chat session once the scheduled interview time has expired.

[1135] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[1136] Step 6: Submit report

[1137] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1138] Example 1

[1139] 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."

[1140] In the traditional recruitment process, companies face the challenge of efficiently managing communication with candidates and interviews. In particular, the enormous amount of time and effort required to process large amounts of candidate data, send appropriately customized scouting emails, and conduct interviews in real time reduces the efficiency of recruitment.

[1141] 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.

[1142] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a specified candidate, means for allowing a user to confirm the content of the scout email, and means for preprocessing input data and converting it into a format suitable for the generative AI model. This automates the creation and sending of scout emails, enabling efficient and effective approaches to candidates.

[1143] The server also includes a means for providing a real-time chat interface and receiving an interview proxy request, a means for analyzing the candidate's responses in real time and generating a next question, a means for analyzing the candidate's response data and generating a report after the interview is completed, a means for sending a notification to the candidate through the real-time chat interface, and a means for storing the interview request in a database, thereby streamlining the entire interview process and facilitating real-time interview handling and subsequent evaluation.

[1144] "Candidate Data" means data containing information about candidates who are evaluated during the recruitment process, including personal information such as name, contact details, work history, and skills.

[1145] A "scouting email template" is a basic format for a scouting email sent to a candidate, and includes placeholders for dynamically inserting candidate information.

[1146] A "generative AI model" is an artificial intelligence model for text generation and natural language generation, which uses deep learning algorithms to generate output text from input data.

[1147] "Individual Information" is specific information extracted from candidate data, such as name, contact details, work history, and skills.

[1148] "Preprocessing" is the process of converting data into a format suitable for generative AI models, and includes operations such as cleaning and normalizing the data.

[1149] The "real-time chat interface" is a chat system that allows candidates and interviewers to communicate in real time, providing an environment where questions and answers can be exchanged instantly.

[1150] An "interview proxy request" is a request made by a user to the system, and includes information necessary to proxy interview a candidate (such as the candidate's name, interview date and time, and position details).

[1151] "Natural language processing technology" is a technology for understanding, analyzing, and generating human language, and is capable of performing processes such as semantic analysis and sentiment analysis of text.

[1152] "Notification" refers to a communication sent by the system to a candidate or user, which may be in the form of email, SMS, etc.

[1153] The present invention is a system that utilizes a generative AI model to automatically generate scouting emails and conduct interviews in real time. Specific embodiments for implementing this system are described below.

[1154] Automatic generation of scout emails

[1155] In this system, the user first inputs candidate data and a scouting email template into the system. Candidate data includes name, contact information, work history, skills, etc. This data is entered using a web form or a dedicated application.

[1156] The server receives this data and stores it in a database. During this process, data integrity is also checked. Next, the server references the stored candidate data and scouting email templates via API and performs data preprocessing. During preprocessing, the data is cleaned, normalized, and converted into a format suitable for generative AI models.

[1157] The preprocessed data is then fed into a generative AI model, such as a Transformer-based natural language generation model (e.g., GPT-3). The generative AI model generates a customized scouting email based on the prompt you provide. For example, it might say, "Hello [candidate name], after reviewing your profile, we believe you would be a great fit for our data scientist position. We would love to speak with you."

[1158] The generated scout email is temporarily stored and an interface is provided for the user to review. Through this interface, the user can check the content of the email and modify it if necessary. Finally, once the user approves the sending of the email, the server sends the approved scout email to the specified candidate. Emails are sent using the SMTP protocol or a cloud-based email API (e.g., SendGrid, Amazon SES).

[1159] Real-time interview proxy

[1160] Next, we will explain the real-time interview proxy process. The user sends an interview proxy request to the system. This request includes the candidate's name, interview date and time, and details of the position. The interview proxy request is sent through a dedicated web application or mobile application.

[1161] The server stores the received interview requests in a database and generates a real-time chat interface using real-time communication technologies such as WebRTC and Socket.io, through which candidates are notified of the interview date and time.

[1162] The candidate accesses the designated chat interface and begins the interview. At the start of the interview, the server selects an initial question from a list of basic questions prepared in advance and sends it to the candidate through the chat interface. For example, a question might be sent: "Hello [candidate's name], thank you for taking the time out of your busy schedule. First, could you tell me about your experience so far?"

[1163] The candidate answers the questions, and the answers are sent to the server. The server uses natural language processing technology to analyze the candidate's responses and evaluates their content. Based on the analysis results, a generative AI model generates the next question and sends it to the candidate in real time. This process is repeated until the end of the interview.

[1164] Once the scheduled interview time is over, the server ends the chat session, analyzes the candidate's response data, and generates a report. This report includes an evaluation of the candidate's response content, tone, and skills. Finally, the generated report is sent to the user. For example, a report may be generated stating, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1165] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1166] Automatic generation of scout emails

[1167] Step 1:

[1168] Input: The user enters candidate data (name, contact details, work history, skills, etc.) and a scouting email template into the system.

[1169] How it works: A user enters candidate data using a web form or dedicated application.

[1170] Output: The entered candidate data and scout email template are sent to the server.

[1171] Step 2:

[1172] Input: The server receives candidate data and a scout email template.

[1173] Operation: The server receives this data and saves it in the database. It checks the data integrity when saving the data.

[1174] Output: Saved candidate data and scouting email template.

[1175] Step 3:

[1176] Input: The server references candidate data and scouting email templates stored in the database.

[1177] How it works: The server retrieves this data through an API, cleans and normalizes it, and converts it into a format suitable for generative AI models.

[1178] Output: Pre-processed candidate data and scout email template.

[1179] Step 4:

[1180] Input: Pre-processed candidate data and scout email template.

[1181] How it works: The server feeds data into a generative AI model (e.g., GPT-3) to generate a customized scouting email based on the prompt text.

[1182] Output: The generated customized scout email.

[1183] Step 5:

[1184] Input: The generated customized scout email.

[1185] Behavior: The server temporarily stores the generated scout email and provides an interface for the user to review it.

[1186] Output: The scout email displayed on the user-facing interface.

[1187] Step 6:

[1188] Input: The user checks the contents of the scout email through the interface.

[1189] How it works: The user can review the contents of the scout email in the interface and make any necessary corrections.

[1190] Output: The corrected or approved Scout email.

[1191] Step 7:

[1192] Input: Approved scout email.

[1193] How it works: The server sends approved scouting emails to the selected candidates using the SMTP protocol or a cloud-based email API (e.g., SendGrid, Amazon SES).

[1194] Output: Scout email sent to candidate.

[1195] Real-time interview proxy

[1196] Step 1:

[1197] Input: The user submits an interview request to the system (candidate name, interview date and time, position details).

[1198] How it works: A user submits a request for an interview through a dedicated web or mobile application.

[1199] Output: The interview request sent to the server.

[1200] Step 2:

[1201] Input: The interview request submitted.

[1202] Operation: The server stores the received interview proxy request in a database.

[1203] Output: Interview requests saved in the database.

[1204] Step 3:

[1205] Input: Interview requests saved in the database.

[1206] How it works: The server generates a real-time chat interface and notifies candidates when their interview is scheduled. The chat interface uses technologies such as WebRTC and Socket.io.

[1207] Output: Interview notice sent to candidate.

[1208] Step 4:

[1209] Input: Candidates who have received interview notices.

[1210] How it works: The candidate accesses the designated chat interface and begins the interview.

[1211] Output: Candidates who accessed the chat interface.

[1212] Step 5:

[1213] Input: The candidate accessing the chat interface.

[1214] How it works: The server selects an initial question from a list of basic questions and sends it to the candidate through a chat interface, such as "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience so far?"

[1215] Output: The initial questions sent to the candidate.

[1216] Step 6:

[1217] Input: Candidate's answers to initial questions.

[1218] How it works: The candidate answers questions and sends their answers to the server.

[1219] Output: Candidate's answers sent to the server.

[1220] Step 7:

[1221] Input: Candidate's answer.

[1222] How it works: The server uses natural language processing techniques to analyze the candidate's responses and evaluate their content.

[1223] Output: Parsed candidate response data.

[1224] Step 8:

[1225] Input: Parsed candidate response data.

[1226] How it works: Based on the analysis results, the server uses a generative AI model to generate the next question and send it to the candidate in real time.

[1227] Output: The next question sent to the candidate.

[1228] Step 9:

[1229] Input: Candidate response data at the end of the interview.

[1230] How it works: Once the allotted interview time has elapsed, the server ends the chat session, analyzes the candidate's response data, and generates a report.

[1231] Output: The generated interview report.

[1232] Step 10:

[1233] Input: The generated interview report.

[1234] Behavior: The server sends the generated report to the user. For example, a report might be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1235] Output: Interview report sent to the user.

[1236] (Application example 1)

[1237] 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."

[1238] In the current recruitment process, creating scouting emails and arranging interviews requires a great deal of effort, placing a heavy burden on companies. Furthermore, as recruitment activities increasingly take place in virtual stores and virtual reality environments, traditional methods are no longer effective. Therefore, a new system is needed to automatically generate scouting emails and efficiently conduct real-time interviews in virtual reality environments.

[1239] 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.

[1240] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a predetermined candidate, means for allowing a user to confirm the content of the scout email, means for displaying the scout email in a virtual reality environment, and means for inputting candidate data and confirming the scout email using a virtual reality head-mounted display, thereby enabling automatic generation of scout emails and effective recruitment activities in a virtual reality environment.

[1241] "Candidate Data" means information about an applicant, including their name, contact details, employment history, skills, and experience.

[1242] A "scout email template" is an email template that contains predefined formats and phrases.

[1243] "Individual Information" is specific details extracted from candidate data.

[1244] A "generative AI model" is an artificial intelligence model that performs transformer-based natural language generation.

[1245] A "customized scouting email" is a custom-made scouting email targeted at a specific candidate, generated based on individual information.

[1246] A "virtual reality environment" is an environment that uses computer graphics and related technologies to immerse a user in a virtual scene or situation.

[1247] A "virtual reality head-mounted display" is a headset-type device that allows a user to visually experience a virtual reality environment.

[1248] A "real-time chat interface" is a user interface for exchanging messages instantly over the Internet.

[1249] An "interview proxy request" is a request from a company or user to the system to conduct an interview on their behalf.

[1250] "Response data" refers to data that includes answers and statements provided by candidates during interviews.

[1251] A "report" is a document that analyzes the results of the interview and includes an overall evaluation and recommendations.

[1252] This invention provides a system for automatically generating scouting emails and conducting real-time interviews using a generative AI model. This system is primarily comprised of a server, a terminal, and a user working together.

[1253] Automatic generation of scout emails

[1254] The server receives candidate data and scouting email templates provided by users and stores them in a database. Candidate data includes names, contact details, work history, skills, and experience. The server then preprocesses this data and converts it into a format suitable for input to a generative AI model. The generative AI model used here is a Transformer-based natural language generation model.

[1255] By inputting the preprocessed data into a generative AI model, an individually customized scout email is generated. For example, a scout email may be automatically generated that reads, "Hello [candidate's name], after looking at your profile, I think you would be a perfect fit for our data scientist position. I would be happy to speak with you." The server temporarily stores this generated scout email for the user and provides an interface for reviewing its contents. Once the user has reviewed and revised it and approved it, the server sends the scout email to the specified candidate.

[1256] Furthermore, the system can display the scout email in a virtual reality environment, allowing users to input candidate data and view the generated scout email using a virtual reality head-mounted display.

[1257] Real-time interview proxy

[1258] The server receives a request for an interview from a user and notifies the candidate of a real-time chat interface at the specified date and time. This chat interface is provided in a virtual reality environment. When the candidate accesses the chat interface, the server selects the first question from a pre-prepared question list and sends it to the candidate.

[1259] For example, a question such as "Hello [candidate's name], thank you for taking the time out of your busy schedule. First, could you tell us about your experience so far?" is sent. When the candidate answers the question, the server analyzes the response in real time, generates the next question, and sends it to the candidate. This is how the interview progresses. When the interview is over, the server analyzes the candidate's response data and generates a report including an overall evaluation. The generated report is sent to the user and includes content such as "The interview results for candidate 'Yamada' are as follows. His experience and skill set are evaluated as being very consistent with our expectations."

[1260] With the above configuration, the system of the present invention can automatically generate scouting emails and conduct effective recruitment activities in a virtual reality environment, thereby improving the efficiency of the entire recruitment process and enabling companies to quickly and effectively recruit talented people.

[1261] Prompt Sentence Examples

[1262] As a concrete example, the following prompt sentence can be considered:

[1263] "Hi [candidate's name], I'm interested in your skill set... I'd love to speak with you."

[1264] In this way, generative AI models and virtual reality environments can be used to recruit more effectively and quickly than traditional methods.

[1265] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1266] Step 1:

[1267] The server receives the candidate data and scouting email template provided by the user and stores it in a database. The input here is the candidate data (name, contact details, work history, skills, etc.) and the scouting email template, and the output is the stored data, which will be used in later processing steps.

[1268] Step 2:

[1269] The server references the stored candidate data and scout email templates and performs preprocessing. Specifically, it cleans and normalizes the data and converts it into a format suitable for the generative AI model. The input is the candidate data and scout email templates, and the output is the preprocessed data. This data is used as input to the generative AI model.

[1270] Step 3:

[1271] The server inputs the preprocessed candidate data and the scout email template into the generative AI model. The generative AI model generates a customized scout email based on the input data. Here, the input is the preprocessed data, and the output is the generated scout email. This scout email contains content that is tailored to each candidate.

[1272] Step 4:

[1273] The server temporarily stores the generated scout email and provides a confirmation interface to the user. The user can check the contents of the scout email through this interface and modify it as necessary. The input is the generated scout email, and the output is the scout email that the user has checked and modified.

[1274] Step 5:

[1275] After the user confirms, modifies, and approves the scout email, the server sends the approved scout email to the specified candidate. The input is the scout email approved by the user, and the output is the sent scout email. This sends the scout email to the candidate.

[1276] Step 6:

[1277] The user uses a virtual reality head-mounted display to input candidate data and view the generated scout email in the virtual reality environment. The input is the scout email displayed in the virtual reality environment, and the output is the scout email viewed by the user.

[1278] Step 7:

[1279] The server receives a request for a job interview from a user and notifies the candidate via a real-time chat interface at the specified date and time. The input is the job interview request and the output is the notification to the candidate. This chat interface is provided within a virtual reality environment.

[1280] Step 8:

[1281] The server manages the interview process through a real-time chat interface. First, it selects initial questions from a pre-prepared question list and sends them to the candidate. The input is the question list, and the output is the questions sent to the candidate.

[1282] Step 9:

[1283] The terminal (candidate) answers questions through a real-time chat interface, and the responses are sent to the server. The input is the candidate's answer, and the output is the response data sent to the server.

[1284] Step 10:

[1285] The server analyzes the candidate's response data and generates the next question. Natural language processing technology is used for the analysis. The input is the candidate's response data, and the output is the generated next question. This question generation is done in real time and is repeated until the end of the interview.

[1286] Step 11:

[1287] Once the interview is over, the server analyzes the candidate's response data and generates a report that includes an overall evaluation of the interview. The input is the response data, and the output is the generated report. This report is sent to the user and includes, for example, "The interview results for candidate 'Yamada' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1288] 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.

[1289] The present invention is a system that recognizes and responds to user emotions in the automatic generation of native-sounding scouting emails and real-time interview proxy by combining an automatic response function using generative AI with an emotion engine. Specific embodiments for implementing this system are described below.

[1290] Automatic generation of scout emails

[1291] 1. Data Acquisition

[1292] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[1293] Server: Receives the data provided by the user and stores it in a database.

[1294] 2. Preparing input data

[1295] Server: Accesses stored candidate data and scouting email templates via API.

[1296] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[1297] 3. Generate scout emails

[1298] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[1299] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[1300] 4. Content adjustment based on sentiment analysis

[1301] Server: Using the emotion engine, the server estimates the emotional state of the potential candidate based on the generated scouting email.

[1302] Server: Optimize the content of the scouting email based on the candidate's estimated emotional state. For example, if the candidate is estimated to be nervous, add more relaxing language.

[1303] 5. Check and send emails

[1304] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify it if necessary.

[1305] User: Check the contents of the scout email and approve sending it.

[1306] Server: Sends approved scouting emails to designated candidates.

[1307] Real-time interview proxy

[1308] 1. Receiving an interview request

[1309] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[1310] Server: Receives interview requests and stores them in a database.

[1311] 2. Preparing the chat interface

[1312] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[1313] Terminal (Candidate): Access the chat interface you were invited to.

[1314] 3. The interview begins

[1315] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[1316] 4. Real-time response processing

[1317] Terminal (candidate): Answer the questions.

[1318] Server: Receives and analyzes the candidate's responses using natural language processing technology to evaluate the content and tone of the applicant's responses.

[1319] Server: Based on the analysis results, the emotion engine estimates the candidate's emotional state. For example, if it estimates that the candidate is nervous, it adjusts the next question to make them feel more relaxed.

[1320] Server: The generative AI model generates the next question.

[1321] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[1322] 5. Interview completion and report generation

[1323] Server: Ends the chat session once the scheduled interview time has expired.

[1324] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[1325] 6. Send report

[1326] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1327] The system of the present invention enables automatic generation of scouting emails and real-time interviews, and by combining this with sentiment analysis, it provides a higher level of personalization and improves the candidate experience, allowing companies to recruit excellent talent quickly and effectively.

[1328] The processing flow will be explained below.

[1329] Automatic generation of scout emails

[1330] Step 1: Data Acquisition

[1331] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[1332] Server: Receives the data provided by the user and stores it in a database.

[1333] Step 2: Prepare input data

[1334] Server: Accesses stored candidate data and scouting email templates via API.

[1335] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, specifically by cleaning and normalizing the data.

[1336] Step 3: Generate a scout email

[1337] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[1338] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and think you'd be a great fit for our position. We'd love to speak with you."

[1339] Step 4: Content adjustment through sentiment analysis

[1340] Server: Uses an emotion engine on the generated scouting email to estimate the emotional state of the potential candidate.

[1341] Server: Optimize the content of the scouting email based on the estimated emotional state. For example, if it is estimated that the candidate is easily nervous, add expressions that will help them relax.

[1342] Step 5: Check and send email

[1343] Server: Temporarily stores the generated scout email for the user to review. Provides an interface that allows the user to review the contents of the email and modify it if necessary.

[1344] User: Check the contents of the scout email and approve sending it.

[1345] Server: Sends approved scouting emails to designated candidates.

[1346] Real-time interview proxy

[1347] Step 1: Receiving an interview request

[1348] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[1349] Server: Receives interview requests and stores them in a database.

[1350] Step 2: Preparing the chat interface

[1351] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[1352] Terminal (Candidate): Access the chat interface you were invited to.

[1353] Step 3: The interview begins

[1354] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, a question might be something like, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[1355] Step 4: Real-time response processing

[1356] Terminal (candidate): Answer the questions.

[1357] Server: Receives and analyzes candidate responses using natural language processing techniques to evaluate the content and tone of the applicant's responses.

[1358] Server: Based on the analysis results, the emotion engine estimates the candidate's emotional state. For example, if the candidate is estimated to be nervous, the next question will be adjusted to make them feel more relaxed.

[1359] Server: The generative AI model generates the next question.

[1360] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[1361] Step 5: End the interview and generate a report

[1362] Server: Ends the chat session once the scheduled interview time has expired.

[1363] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[1364] Step 6: Submit report

[1365] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1366] Example 2

[1367] 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."

[1368] The traditional process of creating scouting emails and conducting interviews required a great deal of time and effort, and it was difficult to properly consider candidates' emotions and reactions. Furthermore, manually creating scouting emails and conducting interviews was inefficient and difficult to respond to individually, preventing the rapid and effective recruitment of excellent talent. Furthermore, the inability to respond to candidates' emotions can lead to a poor candidate experience, putting companies at a disadvantage in the competition for talent.

[1369] 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.

[1370] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for analyzing the generated scout email with an emotion engine and adjusting the content based on the candidate's emotional state, means for sending the adjusted scout email to a specified candidate, and means for allowing a user to confirm the content of the scout email. This enables automatic generation of scout emails and adjustment of the content based on emotions.

[1371] The system also includes a means for providing a real-time chat interface and receiving an interview proxy request, a means for analyzing the candidate's responses in real time and estimating the candidate's emotional state using an emotion engine, a means for using a generative AI model to generate the next question based on the candidate's emotional state, and a means for analyzing the candidate's response data and generating a report after the interview is completed, thereby enabling real-time interview proxy and appropriate responses according to the candidate's emotions.

[1372] "Candidate Data" means data that includes information such as a candidate's name, contact details, work history, and skills.

[1373] A "scout email template" is a template containing a standard email format and structure used to generate scout emails.

[1374] "Generative AI models" are artificial intelligence models that automatically generate text based on input data, including transformer-based natural language generation models.

[1375] An "emotion engine" is an engine that uses text analysis technology to estimate the emotional state of a target and adjusts responses and content based on that.

[1376] A "real-time chat interface" is an interface that enables synchronous communication, and is a system that allows real-time interaction in chat format.

[1377] An "interview proxy request" is a request to the system to proxy an interview, and includes information such as the candidate's name, interview date and time, and position details.

[1378] A "report" is a document generated after the interview is completed that summarizes information including the candidate's response data, sentiment analysis results, and overall evaluation.

[1379] A "software module" is a portion of program code designed to perform a specific function, such as at least partially automatically generating, sentiment analyzing, content adjusting, sending, and reviewing scouting emails.

[1380] The present invention is a system that combines a generative AI model and an emotion engine to automatically generate scouting emails and conduct real-time interviews. Specific embodiments of the present invention are described in detail below.

[1381] Automatic generation of scout emails

[1382] 1. Data Acquisition

[1383] User: The user enters candidate data (such as name, contact details, work history, and skills) and a scouting email template into the system's input form. For example, the user enters "Taro Tanaka" as the candidate's name.

[1384] Server: The server receives the candidate data and email template provided by the user, converts them into JSON format, and stores them in a NoSQL database.

[1385] 2. Preparing input data

[1386] Server: Retrieves stored candidate data and scouting email templates from the database via API. Executes SQL queries to extract the required data.

[1387] Server: Preprocesses the acquired data and converts it into a format suitable for generative AI models. For example, it cleans and normalizes the data, outputs a warning log for missing data, and sets default values.

[1388] 3. Generate scout emails

[1389] Server: Input the preprocessed candidate data and templates into a Transformer-based generative AI model (e.g., GPT-3). The prompt sentence is "Generate a scouting email based on the skill set of candidate 'Taro Tanaka'."

[1390] Server: Receives a customized scouting email generated by the AI ​​model, such as "Hello, Taro Tanaka, after looking at your profile, I think you would be a great fit for our data scientist position. I would be happy to speak with you."

[1391] 4. Content adjustment based on sentiment analysis

[1392] Server: The generated scouting email is sent to the emotion engine, which estimates the candidate's emotional state when they receive it. The content of the email is analyzed using NLP technology.

[1393] Server: Adjust the email's language based on the candidate's estimated emotional state. For example, if the candidate is estimated to be nervous, add a relaxing message.

[1394] 5. Check and send emails

[1395] Server: The generated scout email is temporarily saved and displayed on a dashboard that can be viewed by the user.

[1396] User: The user checks the email on the dashboard, makes any necessary corrections, and once the final check is complete, clicks the send button to approve it.

[1397] Server: Automatically sends approved scouting emails to candidates using the SMTP protocol.

[1398] Real-time interview proxy

[1399] 1. Receiving an interview request

[1400] User: The user enters a request for an interview on the system's dedicated form. The request includes the candidate's name, the scheduled interview date and time, and the desired interview content. For example, the user might enter "Taro Tanaka, interview for a data scientist position, October 1, 2023, 10:00 AM" as the request content.

[1401] Server: The server receives the request data from the user, converts it into JSON format, and saves it in the database.

[1402] 2. Preparing the chat interface

[1403] Server: Generates a real-time chat interface at the specified date and time and sends a notification email to the candidate, including a link to access the interview.

[1404] Device (Candidate): The candidate receives a notification email and clicks on the link to access the chat interface.

[1405] 3. The interview begins

[1406] Server: At the start of the interview, select an initial question from a basic list and send it to the candidate via the chat interface. For example, "Hello, Taro Tanaka. First, please tell me about your experience so far."

[1407] 4. Real-time response processing

[1408] Terminal (candidate): Candidate answers questions through a chat interface.

[1409] Server: Analyzes the received answers in real time and uses an emotion engine to estimate the candidate's emotional state, for example, detecting stress or tension from the tone and content of the answers.

[1410] Server: Generate the next question based on the emotional state. Using a generative AI model, enter the prompt "Generate the next question for the interview with candidate 'Taro Tanaka'."

[1411] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the interview is over.

[1412] 5. Interview completion and report generation

[1413] Server: Ends the chat session once the scheduled interview time has expired.

[1414] Server: Analyzes the interview data and generates a report including an overall evaluation of the interview based on the candidate's response data and sentiment analysis results. For example, it generates a report such as, "The interview results for candidate 'Taro Tanaka' are as follows. His experience and skill set are excellent and meet our expectations."

[1415] 6. Send report

[1416] Server: Automatically sends generated reports to users, either displaying them in a dedicated dashboard or sending them via email.

[1417] In this way, the present invention significantly improves the efficiency of the process of automatically generating scouting emails and conducting real-time interviews, and makes it possible to respond appropriately to the emotions of candidates, thereby enabling companies to recruit excellent talent quickly and effectively.

[1418] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1419] Automatic generation of scout emails

[1420] Step 1: Data Acquisition

[1421] User: Enter candidate data (name, contact details, work history, skills, etc.) and a scouting email template into the input form.

[1422] Input: Candidate data, email template

[1423] Output: User input data

[1424] Specific actions: Enter the required candidate information and template into the system's input form and click the submit button.

[1425] Server: Receives user input data, converts it to JSON format, and saves it in the database.

[1426] Input: User-entered data

[1427] Output: Candidate data stored in database, email template

[1428] Specific operation: Analyzes the received data, converts it into an appropriate format, and stores it in a NoSQL database.

[1429] Step 2: Prepare input data

[1430] Server: Uses API to retrieve candidate data and scout email templates from the database.

[1431] Input: Database query

[1432] Output: Retrieved candidate data, email template

[1433] Specific behavior: Executes SQL queries and extracts stored data.

[1434] Server: Preprocesses the acquired data, complements missing data, and normalizes the data.

[1435] Input: Captured candidate data, email template

[1436] Output: Preprocessed data

[1437] Specific operations: Data is cleaned, normalized, missing data is filled, and the data is converted into a format suitable for generative AI models.

[1438] Step 3: Generate a scout email

[1439] Server: The preprocessed data is input into a generative AI model (e.g., GPT-3) to generate a scouting email.

[1440] Input: Preprocessed data, prompt "Generate a scouting email based on the skill set of candidate 'Taro Tanaka'"

[1441] Output: Generated scout email

[1442] Specific operation: Send data including a prompt sentence to the generative AI model and receive the generated email.

[1443] Step 4: Content adjustment through sentiment analysis

[1444] Server: The generated scouting email is sent to the emotion engine, and the candidate's emotional state upon receiving it is estimated.

[1445] Input: Generated scout email

[1446] Output: Estimated emotional state

[1447] Specific operation: The emotion engine analyzes the content of the email and obtains emotion estimation results.

[1448] Server: Adjusts the wording of the email based on the estimated emotional state.

[1449] Input: Generated scout email, emotion estimation result

[1450] Output: Adjusted scout email

[1451] Specific operation: The content of the email is adjusted according to the emotion estimation results, and relaxing expressions are added.

[1452] Step 5: Check and send email

[1453] Server: Temporarily stores generated and adjusted scout emails and displays them on a dashboard that users can view.

[1454] Input: Adjusted Scout Mail

[1455] Output: Scout email displayed on the dashboard

[1456] Specific behavior: Display the email content in the dashboard interface.

[1457] User: Check the email on the dashboard, make any corrections, and once the final check is complete, click the send button.

[1458] Input: Scout email displayed on the dashboard

[1459] Output: User approved scout email

[1460] Specific actions: Check the email content, make any necessary corrections, and click the send button.

[1461] Server: Sends approved scouting emails to candidates.

[1462] Input: User-approved scout email

[1463] Output: Scout email sent

[1464] What it does: Automatically sends emails using the SMTP protocol.

[1465] Real-time interview proxy

[1466] Step 1: Receiving an interview request

[1467] User: Enters a request for interview on a dedicated form, along with the desired interview date and time and candidate information.

[1468] Input: Interview request data (candidate name, interview date and time, position details, etc.)

[1469] Output: User input data

[1470] Specific actions: Enter the required information in the request form and click the submit button.

[1471] Server: Receives request data from the user and stores it in a database.

[1472] Input: User-entered data

[1473] Output: Request data stored in the database

[1474] Specific operation: Convert the received data into an appropriate format and save it in the database.

[1475] Step 2: Preparing the chat interface

[1476] Server: Generates a real-time chat interface at the specified date and time and sends a notification email to the candidate.

[1477] Input: Request data (interview date and time, candidate contact information)

[1478] Output: Generated chat interface, notification email

[1479] What it does: Generates an email containing a chat interface link and sends it to the candidate.

[1480] Terminal (Candidate): The candidate receives a notification email and accesses the chat interface.

[1481] Input: Notification email, access link

[1482] Output: Chat interface accessed

[1483] Specific action: Click on the link in the email to enter the chat interface.

[1484] Step 3: The interview begins

[1485] Server: At the start of the interview, select initial questions from a basic list and send them to the candidate via a chat interface.

[1486] Input: A list of pre-prepared questions

[1487] Output: Initial question submitted

[1488] What it does: Select an appropriate question from the list of questions and send it to the chat interface. For example, "Hello, Taro Tanaka. First, please tell me about your experience."

[1489] Step 4: Real-time response processing

[1490] Terminal (candidate): Candidate answers questions through a chat interface.

[1491] Input: Initial question

[1492] Output: Candidate's answers

[1493] Specific actions: Enter a response into the chat interface screen and send it.

[1494] Server: Analyzes received answers in real time and uses an emotion engine to estimate the candidate's emotional state.

[1495] Input: Candidate's answer

[1496] Output: Emotional state estimation result

[1497] What it does: It uses natural language processing techniques to analyze responses and infer emotional states, for example detecting stress or tension from the tone and content of responses.

[1498] Server: Generates the next question based on the emotional state.

[1499] Input: Emotional state estimation result, previous answer

[1500] Output: Next question

[1501] Specific operation: The generative AI model is given the prompt "Generate the next question for the interview with candidate 'Taro Tanaka'" and the next question is created.

[1502] Server: Generates the next question and sends it to the candidate in real time.

[1503] Input: Next question

[1504] Output: Question submitted

[1505] What it does: It sends the generated questions to the chat interface and displays them to the candidate. This process is repeated until the interview is over.

[1506] Step 5: End the interview and generate a report

[1507] Server: Ends the chat session once the scheduled interview time has expired.

[1508] Input: Interview End Notice

[1509] Output: Chat sessions completed

[1510] Specific behavior: When the interview end time is reached, a notification will be sent automatically to end the chat session.

[1511] Server: Analyzes the overall interview data and generates a report including an overall evaluation.

[1512] Input: Interview response data, emotion analysis results

[1513] Output: Generated report

[1514] Specific behavior: Create an interview evaluation report based on the candidate's response data and sentiment analysis results. For example, "The interview results for candidate 'Taro Tanaka' are as follows. His experience and skill set match our expectations."

[1515] Step 6: Submit report

[1516] Server: Automatically sends generated reports to users.

[1517] Input: Generated report

[1518] Output: Report sent

[1519] What it does: View the report in a dedicated dashboard or send it via email.

[1520] (Application example 2)

[1521] 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."

[1522] Traditional recruitment and customer service processes have struggled to provide personalized experiences because they struggle to respond appropriately to the emotions of candidates and customers. There is also a need for a way to improve customer satisfaction by understanding customer interests and emotions in real time and generating responses based on them. However, traditional systems lack an efficient means to meet these needs.

[1523] 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.

[1524] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a specified candidate, means for allowing a user to confirm the content of the scout email, means for analyzing the emotional state of a customer in real time via a smart device, means for generating an appropriate response to the customer using the generative AI model based on the analysis results, and means for displaying the generated response on the smart device. This makes it possible to efficiently generate and send personalized responses based on the emotions of candidates and customers.

[1525] "Candidate Data" refers to specific information about people involved in the recruitment process, such as their names, contact details, work history, and skills.

[1526] A "scouting email template" refers to a template that includes the basic structure and sample text of a scouting email to be sent to candidates.

[1527] "Individual Information" refers to specific pieces of information extracted from candidate data.

[1528] A "generative AI model" is an artificial intelligence model that generates natural language based on data, such as a Transformer-based model like GPT-3.

[1529] "Customized scouting email" refers to a scouting email that is tailored to a specific candidate based on the candidate's individual information.

[1530] "Smart devices" refers to devices that can connect to the Internet, such as smartphones, smart glasses, and head-mounted displays.

[1531] "Emotional state" refers to a state that indicates a client's or candidate's psychological reactions and feelings.

[1532] "Analysis results" refers to the results of data processing generated based on emotional state and other input data.

[1533] An "appropriate response" refers to the optimal reply or message that is generated based on the situation and the other person's emotional state.

[1534] "Real-time chat interface" refers to a user interface that enables real-time text and voice communication over the Internet.

[1535] This invention is a system that combines a smart device, a generative AI model, and a sentiment analysis engine to generate personalized responses for candidate and customer interactions. Specific embodiments for implementing this system are described below.

[1536] Hardware and software used

[1537] Hardware: Smartphones, smart glasses, head-mounted displays, cameras

[1538] Software: OpenCV, transformers library (GPT-3 model), EmotionEngine (emotion analysis engine)

[1539] System Program Processing

[1540] The server first receives candidate data and a scout email template and extracts individual information from the candidate data. It then inputs the individual information and the scout email template into a generative AI model to generate a customized scout email. It then sends the generated scout email to the designated candidate and has the user review its content. During this process, the server analyzes the customer's emotional state in real time via the smart device, and generates an appropriate response for the customer using the generative AI model based on the emotion analysis results, which is then displayed on the smart device.

[1541] Specific examples

[1542] For example, a store staff member wearing smart glasses captures a customer's facial expression with a camera and analyzes the image data using Emotion Engine. If Emotion Engine detects that the customer is interested as a result of the analysis, the server sends a prompt to the generative AI model, such as "The customer is interested. Please explain this product to us." The generated response, such as "This product is the latest model, with especially enhanced durability. Please try it out," is displayed in real time on the smart glasses. This allows the store staff member to immediately provide the customer with an appropriate explanation.

[1543] Prompt Sentence Examples

[1544] "The customer is smiling. Can you tell me how to approach them in a relaxed manner?"

[1545] "My customer is looking at my product with interest. What's the best way to approach them and sell them this product?"

[1546] "Customer wants more information. Can you give me a detailed description of this product?"

[1547] The system enables efficient generation and delivery of personalized responses based on candidate and customer sentiment, resulting in a better candidate and customer experience, helping companies hire top talent quickly and effectively and increasing customer satisfaction.

[1548] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1549] Step 1:

[1550] The user enters candidate data and a scouting email template into the system. The entered candidate data includes details such as name, contact information, work history, and skills. This data is sent to the server.

[1551] Step 2:

[1552] The server stores the received candidate data and scouting email templates, makes the stored data available via API, and then performs pre-processing to extract specific information, including cleaning and normalizing the data.

[1553] Step 3:

[1554] The server inputs the preprocessed candidate data and the scouting email template into a generative AI model to generate a customized scouting email. For example, a Transformer-based natural language generation model (GPT-3) is used to generate an email such as, "Hello [candidate name], we've looked at your profile and believe you'd be a great fit for our position. We'd love to speak with you."

[1555] Step 4:

[1556] The server provides a means for the user to check the generated scout email before sending it to a specific candidate. The user checks the contents of the scout email, corrects it if necessary, and then approves its sending.

[1557] Step 5:

[1558] After receiving approval from the user, the server sends a scouting email to the candidate. This sending process is automatic.

[1559] Step 6:

[1560] When a customer visits the store, the device (a user wearing smart glasses or a head-mounted display) captures their facial expressions and movements with a camera, and the captured video data is sent to the server in real time.

[1561] Step 7:

[1562] The server uses the Emotion Engine to analyze the video data sent and determine the customer's emotional state, resulting in an analysis result such as "This customer is showing interest."

[1563] Step 8:

[1564] Based on the results of the emotional state analysis, the server inputs a prompt to the generative AI model, such as "The customer is interested. Please explain this product to me."

[1565] Step 9:

[1566] The generative AI model generates an appropriate response based on the input prompt. The generated response is sent back to the server. For example, a response such as, "This product is the latest model and has been especially strengthened for durability. Please try it out."

[1567] Step 10:

[1568] The server then displays the generated response on smart glasses or a head-mounted display, allowing the user (store staff) to provide appropriate explanations to the customer in real time.

[1569] The input and output at each step is through data processing and data calculations, allowing personalized responses based on candidate or customer sentiment to be automatically generated and sent.

[1570] 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.

[1571] 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.

[1572] 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.

[1573] [Fourth embodiment]

[1574] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1575] 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.

[1576] 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).

[1577] 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.

[1578] 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.

[1579] 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).

[1580] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for 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.

[1581] 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.

[1582] 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.

[1583] 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.

[1584] 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.

[1585] 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.

[1586] 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."

[1587] This invention is a system that utilizes an automatic response function using generative AI to automatically generate native-sounding scouting emails and conduct real-time interviews on behalf of candidates. Specific embodiments for implementing this system are described below.

[1588] Automatic generation of scout emails

[1589] 1. Data Acquisition

[1590] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[1591] Server: Receives the data provided by the user and stores it in a database.

[1592] 2. Preparing input data

[1593] Server: Accesses stored candidate data and scouting email templates via API.

[1594] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[1595] 3. Generate scout emails

[1596] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[1597] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[1598] 4. Sending emails

[1599] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify them if necessary.

[1600] User: Check the contents of the scout email and approve its sending.

[1601] Server: Sends approved scouting emails to the designated candidates.

[1602] Real-time interview proxy

[1603] 1. Receiving an interview request

[1604] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[1605] Server: Receives interview requests and stores them in a database.

[1606] 2. Preparing the chat interface

[1607] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[1608] Terminal (Candidate): Access the chat interface you were invited to.

[1609] 3. The interview begins

[1610] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hi [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[1611] 4. Real-time response processing

[1612] Terminal (candidate): Answer the questions.

[1613] Server: Receives and analyzes the candidate's responses using natural language processing techniques to evaluate the content and tone of the applicant's responses.

[1614] Server: Based on the analysis results, the generative AI model generates the next question.

[1615] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[1616] 5. Interview completion and report generation

[1617] Server: End the chat session once the scheduled interview time is up.

[1618] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[1619] 6. Send report

[1620] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1621] The system of the present invention makes it possible to automatically generate scouting emails and conduct interviews in real time, thereby improving the efficiency of the entire recruitment process and enabling companies to recruit excellent personnel quickly and effectively.

[1622] The processing flow will be explained below.

[1623] Automatic generation of scout emails

[1624] Step 1: Data Acquisition

[1625] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[1626] Server: Receives the data provided by the user and stores it in a database.

[1627] Step 2: Prepare input data

[1628] Server: Accesses stored candidate data and scouting email templates via API.

[1629] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[1630] Step 3: Generate a scout email

[1631] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[1632] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[1633] Step 4: Check and send email

[1634] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify it if necessary.

[1635] User: Check the contents of the scout email and approve sending it.

[1636] Server: Sends approved scouting emails to designated candidates.

[1637] Real-time interview proxy

[1638] Step 1: Receiving an interview request

[1639] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[1640] Server: Receives interview requests and stores them in a database.

[1641] Step 2: Preparing the chat interface

[1642] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[1643] Terminal (Candidate): Access the chat interface you were invited to.

[1644] Step 3: The interview begins

[1645] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[1646] Step 4: Real-time response processing

[1647] Terminal (candidate): Answer the questions.

[1648] Server: Receives and analyzes the candidate's responses using natural language processing technology to evaluate the content and tone of the applicant's responses.

[1649] Server: Based on the analysis results, the generative AI model generates the next question.

[1650] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[1651] Step 5: End the interview and generate a report

[1652] Server: Ends the chat session once the scheduled interview time has expired.

[1653] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[1654] Step 6: Submit report

[1655] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1656] Example 1

[1657] 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."

[1658] In the traditional recruitment process, companies face the challenge of efficiently managing communication with candidates and interviews. In particular, the enormous amount of time and effort required to process large amounts of candidate data, send appropriately customized scouting emails, and conduct interviews in real time reduces the efficiency of recruitment.

[1659] 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.

[1660] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a specified candidate, means for allowing a user to confirm the content of the scout email, and means for preprocessing input data and converting it into a format suitable for the generative AI model. This automates the creation and sending of scout emails, enabling efficient and effective approaches to candidates.

[1661] The server also includes a means for providing a real-time chat interface and receiving an interview proxy request, a means for analyzing the candidate's responses in real time and generating a next question, a means for analyzing the candidate's response data and generating a report after the interview is completed, a means for sending a notification to the candidate through the real-time chat interface, and a means for storing the interview request in a database, thereby streamlining the entire interview process and facilitating real-time interview handling and subsequent evaluation.

[1662] "Candidate Data" means data containing information about candidates who are evaluated during the recruitment process, including personal information such as name, contact details, work history, and skills.

[1663] A "scouting email template" is a basic format for a scouting email sent to a candidate, and includes placeholders for dynamically inserting candidate information.

[1664] A "generative AI model" is an artificial intelligence model for text generation and natural language generation, which uses deep learning algorithms to generate output text from input data.

[1665] "Individual Information" is specific information extracted from candidate data, such as name, contact details, work history, and skills.

[1666] "Preprocessing" is the process of converting data into a format suitable for generative AI models, and includes operations such as cleaning and normalizing the data.

[1667] The "real-time chat interface" is a chat system that allows candidates and interviewers to communicate in real time, providing an environment where questions and answers can be exchanged instantly.

[1668] An "interview proxy request" is a request made by a user to the system, and includes information necessary to proxy interview a candidate (such as the candidate's name, interview date and time, and position details).

[1669] "Natural language processing technology" is a technology for understanding, analyzing, and generating human language, and is capable of performing processes such as semantic analysis and sentiment analysis of text.

[1670] "Notification" refers to a communication sent by the system to a candidate or user, which may be in the form of email, SMS, etc.

[1671] The present invention is a system that utilizes a generative AI model to automatically generate scouting emails and conduct interviews in real time. Specific embodiments for implementing this system are described below.

[1672] Automatic generation of scout emails

[1673] In this system, the user first inputs candidate data and a scouting email template into the system. Candidate data includes name, contact information, work history, skills, etc. This data is entered using a web form or a dedicated application.

[1674] The server receives this data and stores it in a database. During this process, data integrity is also checked. Next, the server references the stored candidate data and scouting email templates via API and performs data preprocessing. During preprocessing, the data is cleaned, normalized, and converted into a format suitable for generative AI models.

[1675] The preprocessed data is then fed into a generative AI model, such as a Transformer-based natural language generation model (e.g., GPT-3). The generative AI model generates a customized scouting email based on the prompt you provide. For example, it might say, "Hello [candidate name], after reviewing your profile, we believe you would be a great fit for our data scientist position. We would love to speak with you."

[1676] The generated scout email is temporarily stored and an interface is provided for the user to review. Through this interface, the user can check the content of the email and modify it if necessary. Finally, once the user approves the sending of the email, the server sends the approved scout email to the specified candidate. Emails are sent using the SMTP protocol or a cloud-based email API (e.g., SendGrid, Amazon SES).

[1677] Real-time interview proxy

[1678] Next, we will explain the real-time interview proxy process. The user sends an interview proxy request to the system. This request includes the candidate's name, interview date and time, and details of the position. The interview proxy request is sent through a dedicated web application or mobile application.

[1679] The server stores the received interview requests in a database and generates a real-time chat interface using real-time communication technologies such as WebRTC and Socket.io, through which candidates are notified of the interview date and time.

[1680] The candidate accesses the designated chat interface and begins the interview. At the start of the interview, the server selects an initial question from a list of basic questions prepared in advance and sends it to the candidate through the chat interface. For example, a question might be sent: "Hello [candidate's name], thank you for taking the time out of your busy schedule. First, could you tell me about your experience so far?"

[1681] The candidate answers the questions, and the answers are sent to the server. The server uses natural language processing technology to analyze the candidate's responses and evaluates their content. Based on the analysis results, a generative AI model generates the next question and sends it to the candidate in real time. This process is repeated until the end of the interview.

[1682] Once the scheduled interview time is over, the server ends the chat session, analyzes the candidate's response data, and generates a report. This report includes an evaluation of the candidate's response content, tone, and skills. Finally, the generated report is sent to the user. For example, a report may be generated stating, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1683] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1684] Automatic generation of scout emails

[1685] Step 1:

[1686] Input: The user enters candidate data (name, contact details, work history, skills, etc.) and a scouting email template into the system.

[1687] How it works: A user enters candidate data using a web form or dedicated application.

[1688] Output: The entered candidate data and scout email template are sent to the server.

[1689] Step 2:

[1690] Input: The server receives candidate data and a scout email template.

[1691] Operation: The server receives this data and saves it in the database. It checks the data integrity when saving the data.

[1692] Output: Saved candidate data and scouting email template.

[1693] Step 3:

[1694] Input: The server references candidate data and scouting email templates stored in the database.

[1695] How it works: The server retrieves this data through an API, cleans and normalizes it, and converts it into a format suitable for generative AI models.

[1696] Output: Pre-processed candidate data and scout email template.

[1697] Step 4:

[1698] Input: Pre-processed candidate data and scout email template.

[1699] How it works: The server feeds data into a generative AI model (e.g., GPT-3) to generate a customized scouting email based on the prompt text.

[1700] Output: The generated customized scout email.

[1701] Step 5:

[1702] Input: The generated customized scout email.

[1703] Behavior: The server temporarily stores the generated scout email and provides an interface for the user to review it.

[1704] Output: The scout email displayed on the user-facing interface.

[1705] Step 6:

[1706] Input: The user checks the contents of the scout email through the interface.

[1707] How it works: The user can review the contents of the scout email in the interface and make any necessary corrections.

[1708] Output: The corrected or approved Scout email.

[1709] Step 7:

[1710] Input: Approved scout email.

[1711] How it works: The server sends approved scouting emails to the selected candidates using the SMTP protocol or a cloud-based email API (e.g., SendGrid, Amazon SES).

[1712] Output: Scout email sent to candidate.

[1713] Real-time interview proxy

[1714] Step 1:

[1715] Input: The user submits an interview request to the system (candidate name, interview date and time, position details).

[1716] How it works: A user submits a request for an interview through a dedicated web or mobile application.

[1717] Output: The interview request sent to the server.

[1718] Step 2:

[1719] Input: The interview request submitted.

[1720] Operation: The server stores the received interview proxy request in a database.

[1721] Output: Interview requests saved in the database.

[1722] Step 3:

[1723] Input: Interview requests saved in the database.

[1724] How it works: The server generates a real-time chat interface and notifies candidates when their interview is scheduled. The chat interface uses technologies such as WebRTC and Socket.io.

[1725] Output: Interview notice sent to candidate.

[1726] Step 4:

[1727] Input: Candidates who have received interview notices.

[1728] How it works: The candidate accesses the designated chat interface and begins the interview.

[1729] Output: Candidates who accessed the chat interface.

[1730] Step 5:

[1731] Input: The candidate accessing the chat interface.

[1732] How it works: The server selects an initial question from a list of basic questions and sends it to the candidate through a chat interface, such as "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience so far?"

[1733] Output: The initial questions sent to the candidate.

[1734] Step 6:

[1735] Input: Candidate's answers to initial questions.

[1736] How it works: The candidate answers questions and sends their answers to the server.

[1737] Output: Candidate's answers sent to the server.

[1738] Step 7:

[1739] Input: Candidate's answer.

[1740] How it works: The server uses natural language processing techniques to analyze the candidate's responses and evaluate their content.

[1741] Output: Parsed candidate response data.

[1742] Step 8:

[1743] Input: Parsed candidate response data.

[1744] How it works: Based on the analysis results, the server uses a generative AI model to generate the next question and send it to the candidate in real time.

[1745] Output: The next question sent to the candidate.

[1746] Step 9:

[1747] Input: Candidate response data at the end of the interview.

[1748] How it works: Once the allotted interview time has elapsed, the server ends the chat session, analyzes the candidate's response data, and generates a report.

[1749] Output: The generated interview report.

[1750] Step 10:

[1751] Input: The generated interview report.

[1752] Behavior: The server sends the generated report to the user. For example, a report might be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1753] Output: Interview report sent to the user.

[1754] (Application example 1)

[1755] 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."

[1756] In the current recruitment process, creating scouting emails and arranging interviews requires a great deal of effort, placing a heavy burden on companies. Furthermore, as recruitment activities increasingly take place in virtual stores and virtual reality environments, traditional methods are no longer effective. Therefore, a new system is needed to automatically generate scouting emails and efficiently conduct real-time interviews in virtual reality environments.

[1757] 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.

[1758] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a predetermined candidate, means for allowing a user to confirm the content of the scout email, means for displaying the scout email in a virtual reality environment, and means for inputting candidate data and confirming the scout email using a virtual reality head-mounted display, thereby enabling automatic generation of scout emails and effective recruitment activities in a virtual reality environment.

[1759] "Candidate Data" means information about an applicant, including their name, contact details, employment history, skills, and experience.

[1760] A "scout email template" is an email template that contains predefined formats and phrases.

[1761] "Individual Information" is specific details extracted from candidate data.

[1762] A "generative AI model" is an artificial intelligence model that performs transformer-based natural language generation.

[1763] A "customized scouting email" is a custom-made scouting email targeted at a specific candidate, generated based on individual information.

[1764] A "virtual reality environment" is an environment that uses computer graphics and related technologies to immerse a user in a virtual scene or situation.

[1765] A "virtual reality head-mounted display" is a headset-type device that allows a user to visually experience a virtual reality environment.

[1766] A "real-time chat interface" is a user interface for exchanging messages instantly over the Internet.

[1767] An "interview proxy request" is a request from a company or user to the system to conduct an interview on their behalf.

[1768] "Response data" refers to data that includes answers and statements provided by candidates during interviews.

[1769] A "report" is a document that analyzes the results of the interview and includes an overall evaluation and recommendations.

[1770] This invention provides a system for automatically generating scouting emails and conducting real-time interviews using a generative AI model. This system is primarily comprised of a server, a terminal, and a user working together.

[1771] Automatic generation of scout emails

[1772] The server receives candidate data and scouting email templates provided by users and stores them in a database. Candidate data includes names, contact details, work history, skills, and experience. The server then preprocesses this data and converts it into a format suitable for input to a generative AI model. The generative AI model used here is a Transformer-based natural language generation model.

[1773] By inputting the preprocessed data into a generative AI model, an individually customized scout email is generated. For example, a scout email may be automatically generated that reads, "Hello [candidate's name], after looking at your profile, I think you would be a perfect fit for our data scientist position. I would be happy to speak with you." The server temporarily stores this generated scout email for the user and provides an interface for reviewing its contents. Once the user has reviewed and revised it and approved it, the server sends the scout email to the specified candidate.

[1774] Furthermore, the system can display the scout email in a virtual reality environment, allowing users to input candidate data and view the generated scout email using a virtual reality head-mounted display.

[1775] Real-time interview proxy

[1776] The server receives a request for an interview from a user and notifies the candidate of a real-time chat interface at the specified date and time. This chat interface is provided in a virtual reality environment. When the candidate accesses the chat interface, the server selects the first question from a pre-prepared question list and sends it to the candidate.

[1777] For example, a question such as "Hello [candidate's name], thank you for taking the time out of your busy schedule. First, could you tell us about your experience so far?" is sent. When the candidate answers the question, the server analyzes the response in real time, generates the next question, and sends it to the candidate. This is how the interview progresses. When the interview is over, the server analyzes the candidate's response data and generates a report including an overall evaluation. The generated report is sent to the user and includes content such as "The interview results for candidate 'Yamada' are as follows. His experience and skill set are evaluated as being very consistent with our expectations."

[1778] With the above configuration, the system of the present invention can automatically generate scouting emails and conduct effective recruitment activities in a virtual reality environment, thereby improving the efficiency of the entire recruitment process and enabling companies to quickly and effectively recruit talented people.

[1779] Prompt Sentence Examples

[1780] As a concrete example, the following prompt sentence can be considered:

[1781] "Hi [candidate's name], I'm interested in your skill set... I'd love to speak with you."

[1782] In this way, generative AI models and virtual reality environments can be used to recruit more effectively and quickly than traditional methods.

[1783] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1784] Step 1:

[1785] The server receives the candidate data and scouting email template provided by the user and stores it in a database. The input here is the candidate data (name, contact details, work history, skills, etc.) and the scouting email template, and the output is the stored data, which will be used in later processing steps.

[1786] Step 2:

[1787] The server references the stored candidate data and scout email templates and performs preprocessing. Specifically, it cleans and normalizes the data and converts it into a format suitable for the generative AI model. The input is the candidate data and scout email templates, and the output is the preprocessed data. This data is used as input to the generative AI model.

[1788] Step 3:

[1789] The server inputs the preprocessed candidate data and the scout email template into the generative AI model. The generative AI model generates a customized scout email based on the input data. Here, the input is the preprocessed data, and the output is the generated scout email. This scout email contains content that is tailored to each candidate.

[1790] Step 4:

[1791] The server temporarily stores the generated scout email and provides a confirmation interface to the user. The user can check the contents of the scout email through this interface and modify it as necessary. The input is the generated scout email, and the output is the scout email that the user has checked and modified.

[1792] Step 5:

[1793] After the user confirms, modifies, and approves the scout email, the server sends the approved scout email to the specified candidate. The input is the scout email approved by the user, and the output is the sent scout email. This sends the scout email to the candidate.

[1794] Step 6:

[1795] The user uses a virtual reality head-mounted display to input candidate data and view the generated scout email in the virtual reality environment. The input is the scout email displayed in the virtual reality environment, and the output is the scout email viewed by the user.

[1796] Step 7:

[1797] The server receives a request for a job interview from a user and notifies the candidate via a real-time chat interface at the specified date and time. The input is the job interview request and the output is the notification to the candidate. This chat interface is provided within a virtual reality environment.

[1798] Step 8:

[1799] The server manages the interview process through a real-time chat interface. First, it selects initial questions from a pre-prepared question list and sends them to the candidate. The input is the question list, and the output is the questions sent to the candidate.

[1800] Step 9:

[1801] The terminal (candidate) answers questions through a real-time chat interface, and the responses are sent to the server. The input is the candidate's answer, and the output is the response data sent to the server.

[1802] Step 10:

[1803] The server analyzes the candidate's response data and generates the next question. Natural language processing technology is used for the analysis. The input is the candidate's response data, and the output is the generated next question. This question generation is done in real time and is repeated until the end of the interview.

[1804] Step 11:

[1805] Once the interview is over, the server analyzes the candidate's response data and generates a report that includes an overall evaluation of the interview. The input is the response data, and the output is the generated report. This report is sent to the user and includes, for example, "The interview results for candidate 'Yamada' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1806] 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.

[1807] The present invention is a system that recognizes and responds to user emotions in the automatic generation of native-sounding scouting emails and real-time interview proxy by combining an automatic response function using generative AI with an emotion engine. Specific embodiments for implementing this system are described below.

[1808] Automatic generation of scout emails

[1809] 1. Data Acquisition

[1810] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[1811] Server: Receives the data provided by the user and stores it in a database.

[1812] 2. Preparing input data

[1813] Server: Accesses stored candidate data and scouting email templates via API.

[1814] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, including cleaning and normalizing the data.

[1815] 3. Generate scout emails

[1816] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[1817] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and believe you'd be a perfect fit for our data scientist position. We'd love to speak with you."

[1818] 4. Content adjustment based on sentiment analysis

[1819] Server: Using the emotion engine, the server estimates the emotional state of the potential candidate based on the generated scouting email.

[1820] Server: Optimize the content of the scouting email based on the candidate's estimated emotional state. For example, if the candidate is estimated to be nervous, add more relaxing language.

[1821] 5. Check and send emails

[1822] Server: The generated scout email is temporarily stored for the user to review. An interface is provided that allows the user to review the contents of the email and modify it if necessary.

[1823] User: Check the contents of the scout email and approve sending it.

[1824] Server: Sends approved scouting emails to designated candidates.

[1825] Real-time interview proxy

[1826] 1. Receiving an interview request

[1827] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[1828] Server: Receives interview requests and stores them in a database.

[1829] 2. Preparing the chat interface

[1830] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[1831] Terminal (Candidate): Access the chat interface you were invited to.

[1832] 3. The interview begins

[1833] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[1834] 4. Real-time response processing

[1835] Terminal (candidate): Answer the questions.

[1836] Server: Receives and analyzes the candidate's responses using natural language processing technology to evaluate the content and tone of the applicant's responses.

[1837] Server: Based on the analysis results, the emotion engine estimates the candidate's emotional state. For example, if it estimates that the candidate is nervous, it adjusts the next question to make them feel more relaxed.

[1838] Server: The generative AI model generates the next question.

[1839] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[1840] 5. Interview completion and report generation

[1841] Server: Ends the chat session once the scheduled interview time has expired.

[1842] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[1843] 6. Send report

[1844] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1845] The system of the present invention enables automatic generation of scouting emails and real-time interviews, and by combining this with sentiment analysis, it provides a higher level of personalization and improves the candidate experience, allowing companies to recruit excellent talent quickly and effectively.

[1846] The processing flow will be explained below.

[1847] Automatic generation of scout emails

[1848] Step 1: Data Acquisition

[1849] User: Enters candidate data and scouting email templates into the system. Candidate data includes name, contact details, work history, skills, etc.

[1850] Server: Receives the data provided by the user and stores it in a database.

[1851] Step 2: Prepare input data

[1852] Server: Accesses stored candidate data and scouting email templates via API.

[1853] Server: Preprocesses the data as needed, converting it into a format suitable for the AI ​​model, specifically by cleaning and normalizing the data.

[1854] Step 3: Generate a scout email

[1855] Server: Inputs the preprocessed candidate data and templates into a generative AI model, which uses a Transformer-based natural language generation model.

[1856] Server: The AI ​​model generates a customized scouting email, such as, "Hi [candidate name], we've looked at your profile and think you'd be a great fit for our position. We'd love to speak with you."

[1857] Step 4: Content adjustment through sentiment analysis

[1858] Server: Uses an emotion engine on the generated scouting email to estimate the emotional state of the potential candidate.

[1859] Server: Optimize the content of the scouting email based on the estimated emotional state. For example, if it is estimated that the candidate is easily nervous, add expressions that will help them relax.

[1860] Step 5: Check and send email

[1861] Server: Temporarily stores the generated scout email for the user to review. Provides an interface that allows the user to review the contents of the email and modify it if necessary.

[1862] User: Check the contents of the scout email and approve sending it.

[1863] Server: Sends approved scouting emails to designated candidates.

[1864] Real-time interview proxy

[1865] Step 1: Receiving an interview request

[1866] User: Sends a request to the system for an interview, including the candidate's name, interview date and time, and details of the position.

[1867] Server: Receives interview requests and stores them in a database.

[1868] Step 2: Preparing the chat interface

[1869] Server: Creates a real-time chat interface and notifies candidates at the specified date and time.

[1870] Terminal (Candidate): Access the chat interface you were invited to.

[1871] Step 3: The interview begins

[1872] Server: At the start of the interview, select an initial question from a list of basic questions prepared in advance and send it to the candidate through a chat interface. For example, a question might be something like, "Hello [candidate name], thank you for taking the time to speak with me. First, can you tell me about your experience?"

[1873] Step 4: Real-time response processing

[1874] Terminal (candidate): Answer the questions.

[1875] Server: Receives and analyzes candidate responses using natural language processing techniques to evaluate the content and tone of the applicant's responses.

[1876] Server: Based on the analysis results, the emotion engine estimates the candidate's emotional state. For example, if the candidate is estimated to be nervous, the next question will be adjusted to make them feel more relaxed.

[1877] Server: The generative AI model generates the next question.

[1878] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the end of the interview.

[1879] Step 5: End the interview and generate a report

[1880] Server: Ends the chat session once the scheduled interview time has expired.

[1881] Server: Analyzes candidate response data and generates a report containing an overall evaluation of the interview.

[1882] Step 6: Submit report

[1883] Server: Sends the generated report to the user. For example, a report may be generated that states, "The interview results for candidate 'Yamada Taro' are as follows. His experience and skill set are evaluated as being highly in line with our expectations."

[1884] Example 2

[1885] 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."

[1886] The traditional process of creating scouting emails and conducting interviews required a great deal of time and effort, and it was difficult to properly consider candidates' emotions and reactions. Furthermore, manually creating scouting emails and conducting interviews was inefficient and difficult to respond to individually, preventing the rapid and effective recruitment of excellent talent. Furthermore, the inability to respond to candidates' emotions can lead to a poor candidate experience, putting companies at a disadvantage in the competition for talent.

[1887] 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.

[1888] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for analyzing the generated scout email with an emotion engine and adjusting the content based on the candidate's emotional state, means for sending the adjusted scout email to a specified candidate, and means for allowing a user to confirm the content of the scout email. This enables automatic generation of scout emails and adjustment of the content based on emotions.

[1889] The system also includes a means for providing a real-time chat interface and receiving an interview proxy request, a means for analyzing the candidate's responses in real time and estimating the candidate's emotional state using an emotion engine, a means for using a generative AI model to generate the next question based on the candidate's emotional state, and a means for analyzing the candidate's response data and generating a report after the interview is completed, thereby enabling real-time interview proxy and appropriate responses according to the candidate's emotions.

[1890] "Candidate Data" means data that includes information such as a candidate's name, contact details, work history, and skills.

[1891] A "scout email template" is a template containing a standard email format and structure used to generate scout emails.

[1892] "Generative AI models" are artificial intelligence models that automatically generate text based on input data, including transformer-based natural language generation models.

[1893] An "emotion engine" is an engine that uses text analysis technology to estimate the emotional state of a target and adjusts responses and content based on that.

[1894] A "real-time chat interface" is an interface that enables synchronous communication, and is a system that allows real-time interaction in chat format.

[1895] An "interview proxy request" is a request to the system to proxy an interview, and includes information such as the candidate's name, interview date and time, and position details.

[1896] A "report" is a document generated after the interview is completed that summarizes information including the candidate's response data, sentiment analysis results, and overall evaluation.

[1897] A "software module" is a portion of program code designed to perform a specific function, such as at least partially automatically generating, sentiment analyzing, content adjusting, sending, and reviewing scouting emails.

[1898] The present invention is a system that combines a generative AI model and an emotion engine to automatically generate scouting emails and conduct real-time interviews. Specific embodiments of the present invention are described in detail below.

[1899] Automatic generation of scout emails

[1900] 1. Data Acquisition

[1901] User: The user enters candidate data (such as name, contact details, work history, and skills) and a scouting email template into the system's input form. For example, the user enters "Taro Tanaka" as the candidate's name.

[1902] Server: The server receives the candidate data and email template provided by the user, converts them into JSON format, and stores them in a NoSQL database.

[1903] 2. Preparing input data

[1904] Server: Retrieves stored candidate data and scouting email templates from the database via API. Executes SQL queries to extract the required data.

[1905] Server: Preprocesses the acquired data and converts it into a format suitable for generative AI models. For example, it cleans and normalizes the data, outputs a warning log for missing data, and sets default values.

[1906] 3. Generate scout emails

[1907] Server: Input the preprocessed candidate data and templates into a Transformer-based generative AI model (e.g., GPT-3). The prompt sentence is "Generate a scouting email based on the skill set of candidate 'Taro Tanaka'."

[1908] Server: Receives a customized scouting email generated by the AI ​​model, such as "Hello, Taro Tanaka, after looking at your profile, I think you would be a great fit for our data scientist position. I would be happy to speak with you."

[1909] 4. Content adjustment based on sentiment analysis

[1910] Server: The generated scouting email is sent to the emotion engine, which estimates the candidate's emotional state when they receive it. The content of the email is analyzed using NLP technology.

[1911] Server: Adjust the email's language based on the candidate's estimated emotional state. For example, if the candidate is estimated to be nervous, add a relaxing message.

[1912] 5. Check and send emails

[1913] Server: The generated scout email is temporarily saved and displayed on a dashboard that can be viewed by the user.

[1914] User: The user checks the email on the dashboard, makes any necessary corrections, and once the final check is complete, clicks the send button to approve it.

[1915] Server: Automatically sends approved scouting emails to candidates using the SMTP protocol.

[1916] Real-time interview proxy

[1917] 1. Receiving an interview request

[1918] User: The user enters a request for an interview on the system's dedicated form. The request includes the candidate's name, the scheduled interview date and time, and the desired interview content. For example, the user might enter "Taro Tanaka, interview for a data scientist position, October 1, 2023, 10:00 AM" as the request content.

[1919] Server: The server receives the request data from the user, converts it into JSON format, and saves it in the database.

[1920] 2. Preparing the chat interface

[1921] Server: Generates a real-time chat interface at the specified date and time and sends a notification email to the candidate, including a link to access the interview.

[1922] Device (Candidate): The candidate receives a notification email and clicks on the link to access the chat interface.

[1923] 3. The interview begins

[1924] Server: At the start of the interview, select an initial question from a basic list and send it to the candidate via the chat interface. For example, "Hello, Taro Tanaka. First, please tell me about your experience so far."

[1925] 4. Real-time response processing

[1926] Terminal (candidate): Candidate answers questions through a chat interface.

[1927] Server: Analyzes the received answers in real time and uses an emotion engine to estimate the candidate's emotional state, for example, detecting stress or tension from the tone and content of the answers.

[1928] Server: Generate the next question based on the emotional state. Using a generative AI model, enter the prompt "Generate the next question for the interview with candidate 'Taro Tanaka'."

[1929] Server: Generates questions and sends them to the candidate in real time. This process is repeated until the interview is over.

[1930] 5. Interview completion and report generation

[1931] Server: Ends the chat session once the scheduled interview time has expired.

[1932] Server: Analyzes the interview data and generates a report including an overall evaluation of the interview based on the candidate's response data and sentiment analysis results. For example, it generates a report such as, "The interview results for candidate 'Taro Tanaka' are as follows. His experience and skill set are excellent and meet our expectations."

[1933] 6. Send report

[1934] Server: Automatically sends generated reports to users, either displaying them in a dedicated dashboard or sending them via email.

[1935] In this way, the present invention significantly improves the efficiency of the process of automatically generating scouting emails and conducting real-time interviews, and makes it possible to respond appropriately to the emotions of candidates, thereby enabling companies to recruit excellent talent quickly and effectively.

[1936] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1937] Automatic generation of scout emails

[1938] Step 1: Data Acquisition

[1939] User: Enter candidate data (name, contact details, work history, skills, etc.) and a scouting email template into the input form.

[1940] Input: Candidate data, email template

[1941] Output: User input data

[1942] Specific actions: Enter the required candidate information and template into the system's input form and click the submit button.

[1943] Server: Receives user input data, converts it to JSON format, and saves it in the database.

[1944] Input: User-entered data

[1945] Output: Candidate data stored in database, email template

[1946] Specific operation: Analyzes the received data, converts it into an appropriate format, and stores it in a NoSQL database.

[1947] Step 2: Prepare input data

[1948] Server: Uses API to retrieve candidate data and scout email templates from the database.

[1949] Input: Database query

[1950] Output: Retrieved candidate data, email template

[1951] Specific behavior: Executes SQL queries and extracts stored data.

[1952] Server: Preprocesses the acquired data, complements missing data, and normalizes the data.

[1953] Input: Captured candidate data, email template

[1954] Output: Preprocessed data

[1955] Specific operations: Data is cleaned, normalized, missing data is filled, and the data is converted into a format suitable for generative AI models.

[1956] Step 3: Generate a scout email

[1957] Server: The preprocessed data is input into a generative AI model (e.g., GPT-3) to generate a scouting email.

[1958] Input: Preprocessed data, prompt "Generate a scouting email based on the skill set of candidate 'Taro Tanaka'"

[1959] Output: Generated scout email

[1960] Specific operation: Send data including a prompt sentence to the generative AI model and receive the generated email.

[1961] Step 4: Content adjustment through sentiment analysis

[1962] Server: The generated scouting email is sent to the emotion engine, and the candidate's emotional state upon receiving it is estimated.

[1963] Input: Generated scout email

[1964] Output: Estimated emotional state

[1965] Specific operation: The emotion engine analyzes the content of the email and obtains emotion estimation results.

[1966] Server: Adjusts the wording of the email based on the estimated emotional state.

[1967] Input: Generated scout email, emotion estimation result

[1968] Output: Adjusted scout email

[1969] Specific operation: The content of the email is adjusted according to the emotion estimation results, and relaxing expressions are added.

[1970] Step 5: Check and send email

[1971] Server: Temporarily stores generated and adjusted scout emails and displays them on a dashboard that users can view.

[1972] Input: Adjusted Scout Mail

[1973] Output: Scout email displayed on the dashboard

[1974] Specific behavior: Display the email content in the dashboard interface.

[1975] User: Check the email on the dashboard, make any corrections, and once the final check is complete, click the send button.

[1976] Input: Scout email displayed on the dashboard

[1977] Output: User approved scout email

[1978] Specific actions: Check the email content, make any necessary corrections, and click the send button.

[1979] Server: Sends approved scouting emails to candidates.

[1980] Input: User-approved scout email

[1981] Output: Scout email sent

[1982] What it does: Automatically sends emails using the SMTP protocol.

[1983] Real-time interview proxy

[1984] Step 1: Receiving an interview request

[1985] User: Enters a request for interview on a dedicated form, along with the desired interview date and time and candidate information.

[1986] Input: Interview request data (candidate name, interview date and time, position details, etc.)

[1987] Output: User input data

[1988] Specific actions: Enter the required information in the request form and click the submit button.

[1989] Server: Receives request data from the user and stores it in a database.

[1990] Input: User-entered data

[1991] Output: Request data stored in the database

[1992] Specific operation: Convert the received data into an appropriate format and save it in the database.

[1993] Step 2: Preparing the chat interface

[1994] Server: Generates a real-time chat interface at the specified date and time and sends a notification email to the candidate.

[1995] Input: Request data (interview date and time, candidate contact information)

[1996] Output: Generated chat interface, notification email

[1997] What it does: Generates an email containing a chat interface link and sends it to the candidate.

[1998] Terminal (Candidate): The candidate receives a notification email and accesses the chat interface.

[1999] Input: Notification email, access link

[2000] Output: Chat interface accessed

[2001] Specific action: Click on the link in the email to enter the chat interface.

[2002] Step 3: The interview begins

[2003] Server: At the start of the interview, select initial questions from a basic list and send them to the candidate via a chat interface.

[2004] Input: A list of pre-prepared questions

[2005] Output: Initial question submitted

[2006] What it does: Select an appropriate question from the list of questions and send it to the chat interface. For example, "Hello, Taro Tanaka. First, please tell me about your experience."

[2007] Step 4: Real-time response processing

[2008] Terminal (candidate): Candidate answers questions through a chat interface.

[2009] Input: Initial question

[2010] Output: Candidate's answers

[2011] Specific actions: Enter a response into the chat interface screen and send it.

[2012] Server: Analyzes received answers in real time and uses an emotion engine to estimate the candidate's emotional state.

[2013] Input: Candidate's answer

[2014] Output: Emotional state estimation result

[2015] What it does: It uses natural language processing techniques to analyze responses and infer emotional states, for example detecting stress or tension from the tone and content of responses.

[2016] Server: Generates the next question based on the emotional state.

[2017] Input: Emotional state estimation result, previous answer

[2018] Output: Next question

[2019] Specific operation: The generative AI model is given the prompt "Generate the next question for the interview with candidate 'Taro Tanaka'" and the next question is created.

[2020] Server: Generates the next question and sends it to the candidate in real time.

[2021] Input: Next question

[2022] Output: Question submitted

[2023] What it does: It sends the generated questions to the chat interface and displays them to the candidate. This process is repeated until the interview is over.

[2024] Step 5: End the interview and generate a report

[2025] Server: Ends the chat session once the scheduled interview time has expired.

[2026] Input: Interview End Notice

[2027] Output: Chat sessions completed

[2028] Specific behavior: When the interview end time is reached, a notification will be sent automatically to end the chat session.

[2029] Server: Analyzes the overall interview data and generates a report including an overall evaluation.

[2030] Input: Interview response data, emotion analysis results

[2031] Output: Generated report

[2032] Specific behavior: Create an interview evaluation report based on the candidate's response data and sentiment analysis results. For example, "The interview results for candidate 'Taro Tanaka' are as follows. His experience and skill set match our expectations."

[2033] Step 6: Submit report

[2034] Server: Automatically sends generated reports to users.

[2035] Input: Generated report

[2036] Output: Report sent

[2037] What it does: View the report in a dedicated dashboard or send it via email.

[2038] (Application example 2)

[2039] 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."

[2040] Traditional recruitment and customer service processes have struggled to provide personalized experiences because they struggle to respond appropriately to the emotions of candidates and customers. There is also a need for a way to improve customer satisfaction by understanding customer interests and emotions in real time and generating responses based on them. However, traditional systems lack an efficient means to meet these needs.

[2041] 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.

[2042] In this invention, the server includes means for receiving candidate data and a scout email template and extracting individual information from the candidate data, means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email, means for sending the generated scout email to a specified candidate, means for allowing a user to confirm the content of the scout email, means for analyzing the emotional state of a customer in real time via a smart device, means for generating an appropriate response to the customer using the generative AI model based on the analysis results, and means for displaying the generated response on the smart device. This makes it possible to efficiently generate and send personalized responses based on the emotions of candidates and customers.

[2043] "Candidate Data" refers to specific information about people involved in the recruitment process, such as their names, contact details, work history, and skills.

[2044] A "scouting email template" refers to a template that includes the basic structure and sample text of a scouting email to be sent to candidates.

[2045] "Individual Information" refers to specific pieces of information extracted from candidate data.

[2046] A "generative AI model" is an artificial intelligence model that generates natural language based on data, such as a Transformer-based model like GPT-3.

[2047] "Customized scouting email" refers to a scouting email that is tailored to a specific candidate based on the candidate's individual information.

[2048] "Smart devices" refers to devices that can connect to the Internet, such as smartphones, smart glasses, and head-mounted displays.

[2049] "Emotional state" refers to a state that indicates a client's or candidate's psychological reactions and feelings.

[2050] "Analysis results" refers to the results of data processing generated based on emotional state and other input data.

[2051] An "appropriate response" refers to the optimal reply or message that is generated based on the situation and the other person's emotional state.

[2052] "Real-time chat interface" refers to a user interface that enables real-time text and voice communication over the Internet.

[2053] This invention is a system that combines a smart device, a generative AI model, and a sentiment analysis engine to generate personalized responses for candidate and customer interactions. Specific embodiments for implementing this system are described below.

[2054] Hardware and software used

[2055] Hardware: Smartphones, smart glasses, head-mounted displays, cameras

[2056] Software: OpenCV, transformers library (GPT-3 model), EmotionEngine (emotion analysis engine)

[2057] System Program Processing

[2058] The server first receives candidate data and a scout email template and extracts individual information from the candidate data. It then inputs the individual information and the scout email template into a generative AI model to generate a customized scout email. It then sends the generated scout email to the designated candidate and has the user review its content. During this process, the server analyzes the customer's emotional state in real time via the smart device, and generates an appropriate response for the customer using the generative AI model based on the emotion analysis results, which is then displayed on the smart device.

[2059] Specific examples

[2060] For example, a store staff member wearing smart glasses captures a customer's facial expression with a camera and analyzes the image data using Emotion Engine. If Emotion Engine detects that the customer is interested as a result of the analysis, the server sends a prompt to the generative AI model, such as "The customer is interested. Please explain this product to us." The generated response, such as "This product is the latest model, with especially enhanced durability. Please try it out," is displayed in real time on the smart glasses. This allows the store staff member to immediately provide the customer with an appropriate explanation.

[2061] Prompt Sentence Examples

[2062] "The customer is smiling. Can you tell me how to approach them in a relaxed manner?"

[2063] "My customer is looking at my product with interest. What's the best way to approach them and sell them this product?"

[2064] "Customer wants more information. Can you give me a detailed description of this product?"

[2065] The system enables efficient generation and delivery of personalized responses based on candidate and customer sentiment, resulting in a better candidate and customer experience, helping companies hire top talent quickly and effectively and increasing customer satisfaction.

[2066] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2067] Step 1:

[2068] The user enters candidate data and a scouting email template into the system. The entered candidate data includes details such as name, contact information, work history, and skills. This data is sent to the server.

[2069] Step 2:

[2070] The server stores the received candidate data and scouting email templates, makes the stored data available via API, and then performs pre-processing to extract specific information, including cleaning and normalizing the data.

[2071] Step 3:

[2072] The server inputs the preprocessed candidate data and the scouting email template into a generative AI model to generate a customized scouting email. For example, a Transformer-based natural language generation model (GPT-3) is used to generate an email such as, "Hello [candidate name], we've looked at your profile and believe you'd be a great fit for our position. We'...

Claims

1. a means for receiving candidate data and a scout email template and extracting individual information from the candidate data; A means for inputting the individual information and the scout email template into a generative AI model to generate a customized scout email; means for sending the generated scout email to a predetermined candidate; means for allowing the user to confirm the contents of the scout email; A system including:

2. means for providing a real-time chat interface and receiving interview substitution requests; means for analyzing the candidate's response in real time and generating a next question; A means of analyzing candidate response data and generating reports after the interview is completed; The system of claim 1 , comprising:

3. The system according to claim 1 , further comprising a software module that performs at least a part of the automatic generation, transmission and confirmation of the scout email.

Images