Information processing systems, information processing methods, and programs

Abstract

To efficiently improve negotiation skills in sales activities and provide role-playing training that combines diverse speaker attributes and diverse negotiation scenarios. [Solution] An information processing system that supports sales role-playing using a generating AI, comprising: a response attribute template storage unit that stores multiple response attribute templates that categorize the way the other party speaks, etc.; a scene behavior template storage unit that stores multiple scene behavior templates, including questions, concerns, counterarguments, or disclosure information that the other party should present during the negotiation phase, etc., separately from the response attribute templates; a reception unit that receives response attribute identification information and scene identification information; and a prompt generation unit that generates prompts that instruct the generating AI to combine the identified response attribute templates and scene behavior templates and behave as a party that speaks in accordance with the response attribute templates and asks questions, etc., in accordance with the scene behavior templates.

Description

Technical Field

[0001] The present invention relates to an information processing system, an information processing method, and a program.

Background Art

[0002] A system for providing a data management service on the Internet is known (see Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In a conventional enterprise legacy system, a connection is made from a personal computer owned by a customer via a dial-up connection, and custom software on the customer workstation is required.

[0005] The present invention has been made in view of such a background, and an object thereof is to provide a technology capable of efficiently improving negotiation skills in business activities.

Means for Solving the Problems

[0006] The main invention of the present invention for solving the above problems is an information processing system that supports sales role-playing using generating AI, comprising: a response attribute template storage unit that stores a plurality of response attribute templates that categorize personal attribute or conversation attribute characteristics including at least one of the speaking style, attitude, questioning tendency or decision-making tendency of the other party in the role-playing; and a scene behavior template storage unit that stores a plurality of scene behavior templates corresponding to the negotiation phase or negotiation objective, and which include at least one of the questions, concerns, objections, requests, disclosure information, non-disclosure information, progress conditions or termination conditions that the other party should present in the negotiation phase or negotiation objective, separately from the response attribute templates, and The system is characterized by comprising: a reception unit that receives response attribute identification information that identifies one response attribute template from among multiple response attribute templates, and scene identification information that identifies one scene action template from among multiple scene action templates; and a prompt generation unit that combines the response attribute template identified by the response attribute identification information and the scene action template identified by the scene identification information to generate a prompt that instructs the generating AI to behave as a role-playing partner, speaking or acting in accordance with the response attribute template and asking questions, expressing concerns, raising objections, making requests, or disclosing information in accordance with the scene action template.

[0007] Further issues and solutions disclosed in this application will be made clear in the section on embodiments of the invention and in the drawings. [Effects of the Invention]

[0008] According to the present invention, negotiation skills in sales activities can be efficiently improved. [Brief explanation of the drawing]

[0009] [Figure 1] This figure shows an example of the overall configuration of an information processing system. [Figure 2] This figure shows an example of the hardware configuration of management server 2. [Figure 3]This figure shows an example of the software configuration for management server 2. [Figure 4] This figure shows an example of the software configuration of user terminal 1. [Figure 5] This is a diagram illustrating the processing flow in an information processing system. [Figure 6] This figure shows an example of the data structure for response attribute templates and scene behavior templates. [Figure 7] This diagram shows a configuration that generates prompts by combining response attribute templates and scene behavior templates. [Modes for carrying out the invention]

[0010] <System Overview> The following describes an information processing system according to one embodiment of the present invention. The information processing system of this embodiment is a system that supports the improvement of sales representatives' negotiation skills using generative AI (generative artificial intelligence). Specifically, it allows sales representatives to perform role-playing assuming a specific sales partner before an actual negotiation. The system sets an appropriate persona based on the unique information of the sales partner, and the generative AI acts as that persona to provide a realistic negotiation simulation. Furthermore, after the role-playing is completed, an evaluation function can analyze the sales representative's performance and suggest areas for improvement. In addition, the information processing system of this embodiment separately stores response attribute templates that categorize personal attribute characteristics such as the other party's way of speaking and attitude, and scene behavior templates that define the typical behavior of the other party in the negotiation phase. By independently selecting and combining these, it is possible to provide role-playing that combines diverse speaker attributes and diverse negotiation scenes, regardless of whether or not there is unique information of a specific sales partner.

[0011] Figure 1 shows an example of the overall configuration of an information processing system. The information processing system in this embodiment includes a management server 2. The management server 2 is connected to the user terminal 1 via a communication network. The communication network is, for example, the internet, and is constructed using public telephone lines, mobile phone lines, wireless communication channels, Ethernet (registered trademark), etc. The management server 2 is also connected to the sales management DB (database) 3, the business negotiation management DB 4, and the document DB 5 via communication.

[0012] User terminal 1 is a computer operated by a user who is a sales representative. User terminal 1 can be, for example, a smartphone, tablet computer, or personal computer. Through user terminal 1, the user can select a sales client to role-play with and conduct role-playing using text or voice.

[0013] Management Server 2 may be a general-purpose computer such as a workstation or personal computer, or it may be logically implemented through cloud computing. Management Server 2 utilizes a generative AI engine, either internally or as an external service, to enable role-playing with the user.

[0014] <Management Server> Figure 2 shows an example of the hardware configuration of the management server 2. Note that the illustrated configuration is just one example, and other configurations are also possible. The management server 2 includes a CPU 201, memory 202, storage device 203, communication interface 204, input device 205, and output device 206. The storage device 203 stores various data and programs, such as a hard disk drive, solid-state drive, or flash memory. The communication interface 204 is an interface for connecting to a communication network, such as an adapter for connecting to Ethernet®, a modem for connecting to a public telephone network, a wireless communication device for wireless communication, or a USB (Universal Serial Bus) connector or RS232C connector for serial communication. The input device 205 is for inputting data, such as a keyboard, mouse, touch panel, button, or microphone. The output device 206 is for outputting data, such as a display, printer, or speaker. Furthermore, each functional unit of the management server 2, as described later, is realized by the CPU 201 reading programs stored in the storage device 203 into memory 202 and executing them, and each storage unit of the management server 2 is realized as part of the storage area provided by memory 202 and storage device 203.

[0015] Figure 3 shows an example of the software configuration of the management server 2. The management server 2 includes a sales partner storage unit 231, a persona storage unit 232, a sales partner reception unit 211, a persona setting unit 212, a role-playing reception unit 213, a response generation unit 214, an output unit 215, a unique information acquisition unit 216, a business negotiation information acquisition unit 217, a sales material acquisition unit 218, an evaluation unit 219, and a candidate selection UI unit 220. Furthermore, as shown in Figure 3, the management server 2 also includes a response attribute template storage unit 233, a scene behavior template storage unit 234, a response attribute reception unit 221, a scene reception unit 222, and a prompt generation unit 223.

[0016] Figure 4 shows an example of the software configuration of user terminal 1. User terminal 1 comprises a display unit 111, an input unit 112, a communication unit 113, and a storage unit 131.

[0017] <Management Server 2> The functions of Management Server 2 will be described below.

[0018] The business partner storage unit 231 stores the unique information of a business partner in association with the business partner. The unique information includes, for example, the company name of the business partner, industry type, scale, location, main products / services, business policy, organizational structure, URL of the website, main trading partners, and positioning within the industry. These pieces of information are stored in association with partner identification information (such as customer ID, etc.) that can be uniquely identified for each business partner.

[0019] The persona storage unit 232 stores a plurality of personas of the role-playing partner. A persona is a fictional character image that defines the attributes, personality, behavior patterns, values, decision-making criteria, etc. of that character. In this embodiment, various types of character images that may be encountered in business activities, such as "conservative manager", "innovative IT department head", "cost-conscious purchasing staff", "technology-oriented engineer", etc., are defined as personas. Each persona is associated with a persona ID, a persona name, and a profile that describes the detailed characteristics of the persona. Note that in this specification, the attributes, personality, behavior patterns, values, or decision-making criteria defined by a persona refer to personal attribute or conversation attribute characteristics such as speaking style, attitude, reaction tendency, or decision-making tendency, and do not include scene-dependent behaviors such as questions, concerns, counterarguments, disclosed information, or progress conditions that the other party should present in a specific negotiation phase. Such scene-dependent behaviors are defined by the scene action template described later.

[0020] The Sales Partner Reception Unit 211 accepts the designation of a sales partner. Specifically, it receives the contact identification information of the sales partner specified by the user via communication from the user terminal 1. The Sales Partner Reception Unit 211 provides the received contact identification information to the Persona Setting Unit 212. The Sales Partner Reception Unit 211 can also accept the designation of a sales partner indirectly. For example, when a user is viewing specific deal information, and a form such as a button for role-playing related to that deal information is selected, the sales partner related to that deal information can be considered to have been selected.

[0021] The persona setting unit 212 identifies a persona corresponding to a specified sales target and provides the generating AI with a prompt instructing it to behave in a manner similar to that of the identified persona. Specifically, the persona setting unit 212 retrieves corresponding unique information from the sales target memory unit 231 based on the contact identification information received from the sales target reception unit 211. Next, based on the retrieved unique information (e.g., industry and size), it selects an appropriate persona from the persona memory unit 232. For example, for a small or medium-sized manufacturing company, it would select the "cost-conscious manager" persona, and for a large IT company, it would select the "innovative IT department head" persona. The persona setting unit 212 combines the profile of the selected persona with the unique information of the sales target to create a prompt for the generating AI. This prompt includes instructions such as, "You are the head of the □□ department at △△ company in the 〇〇 industry. You have the characteristic of (persona characteristic). Please respond in the following conversation based on these characteristics."

[0022] The role-playing reception unit 213 receives input for role-playing from the user. Specifically, it receives text and voice data entered by the user via the user terminal 1 and converts the voice data to text as needed. The role-playing reception unit 213 provides the received input to the response generation unit 214.

[0023] The response generation unit 214 generates a response by providing the generating AI with prompts that include inputs and instructions for generating responses to those inputs. Specifically, the response generation unit 214 sends a prompt to the generating AI that includes user input received from the role-playing reception unit 213 and persona information set by the persona setting unit 212. Based on this information, the generating AI generates an appropriate response for the specified persona. For example, if the persona is a "cost-conscious manager," there will be many questions about price and payback period, and if the persona is a "technology-oriented engineer," there will be many questions about technical details and compatibility, thus generating responses that reflect the characteristics of the persona. The response generation unit 214 provides the response received from the generating AI to the output unit 215.

[0024] The output unit 215 outputs a response to the user. Specifically, it sends the response text received from the response generation unit 214 to the user terminal 1. Alternatively, the response text can be converted into audio data using the text-to-speech conversion function and sent to the user terminal 1.

[0025] The unique information acquisition unit 216 acquires unique information about sales targets from an externally installed sales management database 3. Specifically, based on the target identification information received by the sales target reception unit 211, it issues a query to the sales management DB 3 to obtain detailed information about the corresponding sales target. The acquired information is stored in the sales target storage unit 231 and also provided to the persona setting unit 212. In addition, the unique information acquisition unit 216 can also acquire the web page of an organization using the URL of the organization included in the unique information and provide the acquired web page to the generation AI. This allows the generation AI to learn the latest information and official expressions of sales targets, enabling more realistic role-playing.

[0026] The sales opportunity information acquisition unit 217 uses the designated sales partner as a key to acquire past sales opportunity information from the external sales opportunity management database 4 and provides this sales opportunity information to the generating AI. Specifically, based on the partner identification information received by the sales partner reception unit 211, it issues a query to the sales opportunity management DB 4 to acquire information such as past sales opportunity records, proposal details, customer questions and concerns, and the outcome of whether a deal was closed or not. The acquired information is added to the prompts created by the persona setting unit 212 and provided to the generating AI. This enables more realistic role-playing based on past sales opportunity history.

[0027] The sales material acquisition unit 218 acquires sales materials from an external material database 5 based on material identification information associated with the user, and provides these sales materials to the generation AI. Specifically, it acquires relevant sales materials such as product catalogs, proposal templates, price lists, and technical specifications from the material DB 5 based on information such as the user's ID, department, and assigned products. The acquired materials are then provided to the generation AI after undergoing text extraction processing, etc. This allows the generation AI to generate responses based on actual product and pricing information.

[0028] The evaluation unit 219 acquires inputs and responses, and after the responses meet predetermined termination conditions, it obtains an evaluation result for the entire role-play from the generating AI and outputs it to the user. Specifically, it records the entire process of the role-play (history of user inputs and generating AI responses), and when the role-play ends, it gives the generating AI another prompt to request an evaluation. This prompt includes instructions such as, "Analyze the following sales role-play conversation and point out its strengths, weaknesses, and areas for improvement." The evaluation unit 219 can also refer to sales materials acquired by the sales material acquisition unit 218, calculate the accuracy of the user's statements, and include this in the evaluation result. For example, it can verify whether the content of the user's explanation of product features and price matches the content of the actual sales materials and calculate an accuracy score.

[0029] The candidate selection UI unit 220 displays a list of multiple potential sales partners to the user and provides the recipient identification information of the selected sales partner to the reception unit. Specifically, it displays a list of sales partners obtained from the sales management DB 3 on the user terminal 1 and provides the information of the sales partner selected by the user to the sales partner reception unit 211. The candidate selection UI unit 220 also provides filtering functions such as industry, size, and region of sales partners, as well as keyword search functions, to enable the user to efficiently select the desired sales partner.

[0030] <User Terminal 1> The following describes the functional components of user terminal 1.

[0031] The memory unit 131 is the storage device of the user terminal 1. The memory unit 131 stores user information, role-playing history, sales material cache, etc. User information includes user ID, name, department, assigned product, access rights, etc.

[0032] The display unit 111 displays the user interface. Specifically, it displays the sales partner selection screen, the role-playing conversation screen, the evaluation results display screen, and so on. The display unit 111 configures the screen based on the data received from the management server 2 and presents it to the user.

[0033] The input unit 112 accepts input from the user. Specifically, it accepts text input, touch operations, voice input, etc. During role-playing, it accepts the user's statements as text or voice and transmits them to the management server 2 via the communication unit 113.

[0034] The communication unit 113 communicates with the management server 2. Specifically, it transmits user input received by the input unit 112 to the management server 2 and provides the response data received from the management server 2 to the display unit 111. The communication unit 113 uses protocols such as HTTP (Hypertext Transfer Protocol) and WebSocket for communication.

[0035] Figure 5 is a diagram illustrating the processing flow in an information processing system.

[0036] First, the user operating user terminal 1 selects a sales partner via the candidate selection UI unit 220 (S501). Information on the selected sales partner is transmitted to the sales partner reception unit 211 (S502). Next, the unique information acquisition unit 216 acquires unique information of the selected sales partner from the sales management DB 3 (S503). In addition, the business negotiation information acquisition unit 217 acquires past business negotiation information for the same sales partner from the business negotiation management DB 4 (S504). Furthermore, the sales material acquisition unit 218 acquires sales materials related to the user from the material DB 5 (S505).

[0037] After acquiring this information, the persona setting unit 212 selects an appropriate persona from the persona storage unit 232 based on the unique information of the sales partner (S506), and combines the selected persona with the acquired information to create a prompt for the generating AI (S507).

[0038] When role-playing begins, the user inputs a statement via the input unit 112 of the user terminal 1 (S508), and the input statement is sent to the role-playing reception unit 213 (S509). The response generation unit 214 sends the user's statement and the set prompt to the generation AI (S510), and obtains a response from the generation AI (S511). The obtained response is sent to the user terminal 1 via the output unit 215 (S512), and is displayed on the display unit 111 of the user terminal 1 (S513).

[0039] This role-playing exchange (S508-S513) is repeated until the termination condition is met (S514). When the termination condition is met, the evaluation unit 219 sends the entire role-playing history to the generating AI (S515) and obtains the evaluation result (S516). The obtained evaluation result is sent to the user terminal 1 (S517) and displayed to the user (S518).

[0040] As described above, the information processing system of this embodiment allows sales representatives to conduct role-playing exercises with specific sales targets before actual business negotiations, utilizing the generation AI. An appropriate persona is set based on the unique information of the sales target, and the generation AI acts as that persona, realizing a realistic negotiation simulation. Furthermore, by utilizing past negotiation information and sales materials, more practical training becomes possible. In addition, after the role-playing is completed, the evaluation function can analyze the sales representative's performance and suggest areas for improvement. As a result, sales representatives can efficiently improve their negotiation skills, and an increase in the success rate of actual sales activities can be expected.

[0041] <Separation of response attribute templates and scene action templates> The information processing system of this embodiment stores, as separate data, response attribute templates that categorize the personal or conversational characteristics of the other party, and scene behavior templates that define the typical behavior of the other party in a business negotiation scenario, as data for controlling the other party in a role-playing scenario. The response attribute templates mainly define "how the other party speaks," and the scene behavior templates mainly define "what the other party asks, what concerns they express, and under what conditions they agree or disagree" in that business negotiation scenario. In other words, personal attribute elements such as speaking style and scene-dependent elements such as behavior in a business negotiation scenario are stored separately in separate templates.

[0042] A response attribute template is a template that categorizes personal or conversational attribute characteristics of the other party in a role-playing scenario, such as their speaking style, tone of voice, volume of speech, questioning tendencies, emotional expression, attitude, caution, or decision-making tendencies. It does not require specific identifying information of the other party, such as the name, company name, or URL of a particular sales partner. A response attribute template can be stored as categorical information that can be applied to multiple sales partners. Furthermore, a response attribute template (including the part that defines personal attribute characteristics when simply referred to as a persona in this specification) does not include scene-dependent behaviors such as questions, concerns, objections, disclosed information, or conditions for progress that the other party should present during the negotiation phase; these are defined by the scene behavior template.

[0043] A scene behavior template is a template that defines the negotiation phase, negotiation objective, the role of the other party in that scene, the questions, concerns, objections, demands, information to be disclosed, information not to be disclosed, conditions for transitioning to the next phase, termination conditions, or evaluation criteria, and is stored separately from the response attribute template. While the response attribute template mainly defines the way the other party speaks or their attitude, the scene behavior template mainly defines what questions the other party will ask, what concerns they will express, and under what conditions they will disclose information or move the negotiation phase forward in that negotiation scene.

[0044] Figure 6 shows an example of the data structure for response attribute templates and scene behavior templates. A response attribute template stores, for example, a response attribute template ID that identifies the response attribute template, a type name, speaking style, speaking volume, questioning tendency, decision-making tendency, and emotional expression, all associated with each other. For example, the response attribute template ID "P01" is associated with the type name "cautious / reserved," speaking style "short, states conclusion first," speaking volume "little," questioning tendency "confirms evidence, implementation track record, and risks," decision-making tendency "prioritizes failure avoidance," and emotional expression "difficult to express emotions." Response attribute templates do not include names, company names, or URLs that identify specific sales clients.

[0045] As shown in Figure 6, a scene action template stores, for example, a scene template ID that identifies the scene action template, the negotiation phase, the scene objective, the role of the other party in the scene, initial disclosure information, typical questions, typical counterarguments, information disclosure conditions, conditions for transitioning to the next phase, and evaluation items. For example, the scene template ID "S02" is associated with the negotiation phase "Initial hearing", the scene objective "Training to see if we can elicit the issues", the role of the other party in the scene "Issues holder in the field department", initial disclosure information "Explaining the issues vaguely", typical questions "Confirming the effects of implementation, results from other companies, and cost estimates", typical counterarguments "There are no major problems with the current operation", information disclosure conditions "Disclose specific issues if the user asks follow-up questions two or more times", conditions for transitioning to the next phase "When issues, budget, and decision-makers have been confirmed", and evaluation items "In-depth questioning, hypothesis presentation, agreement at next meeting".

[0046] The response attribute template storage unit 233 stores the above-mentioned multiple response attribute templates. The scene action template storage unit 234 stores the above-mentioned multiple scene action templates separately from the response attribute templates.

[0047] The response attribute receiving unit 221 receives response attribute identification information that identifies one response attribute template from among multiple response attribute templates. The scene receiving unit 222 receives scene identification information that identifies one scene action template from among multiple scene action templates, separately from the response attribute identification information. The response attribute identification information and the scene identification information are received, for example, through a user interface having a first selection field for selecting a response attribute template and a second selection field for selecting a scene action template.

[0048] The prompt generation unit 223 generates prompts for the generating AI by combining the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information. These prompts include instructions to behave as a role-playing partner in a manner or attitude that conforms to the response attribute template, and instructions to ask questions, express concerns, raise objections, make requests, or disclose information in accordance with the scene behavior template. For example, the prompt generation unit 223 might give the generating AI instructions such as, "You are a partner who speaks in a manner / attitude (from the response attribute template). The current negotiation phase is (from the scene behavior template), and you should behave according to (typical questions, typical objections, and information disclosure conditions from the scene behavior template)."

[0049] The response generation unit 214 generates a response by providing the prompt generated by the prompt generation unit 223 and the user's input to the generating AI, and the output unit 215 outputs the response to the user. The prompt generation unit 223 also updates the scene state included in the scene action template according to the conversation history during the role-playing. The scene state includes states such as "issue not disclosed," "issue partially disclosed," "budget confirmed," and "next meeting agreed," and the generating AI changes the questions, counterarguments, or disclosed information that can be presented according to the scene state. This enables realistic role-playing in which information is disclosed or the phases progress in stages according to the user's responses.

[0050] The evaluation unit 219 compares the evaluation items or progress conditions included in the scene action template with the user's input history to evaluate whether the user has performed the necessary confirmations, explanations, or agreements in the relevant negotiation phase, and outputs the evaluation results to the user. For example, if the initial hearing scene action template is selected, the evaluation unit 219 evaluates whether the user has deepened the discussion of the issue, presented hypotheses, and reached an agreement for the next meeting, based on the evaluation items of the scene action template.

[0051] Figure 7 shows a configuration for generating prompts by combining a response attribute template and a scene behavior template. An example of the processing flow in this configuration is described below. First, the user or system selects a response attribute template. Next, the user or system selects a scene behavior template independently of the response attribute template. The prompt generation unit 223 combines the selected response attribute template and scene behavior template and instructs the generating AI to control the way of speaking, attitude, and reaction tendencies according to the response attribute template, and to control questions, counterarguments, information disclosure, and phase progression according to the scene behavior template. During role-playing, the scene state in the scene behavior template is updated based on the conversation history. After the role-playing is completed, the user's utterances are evaluated based on the evaluation items included in the scene behavior template.

[0052] For example, the user selects "cautious / reserved" as the response attribute template and "price negotiation phase" as the scene action template. In this case, the prompt generation unit 223 instructs the generating AI to speak cautiously, confirming the basis with short utterances, while also asking questions about typical counterparty behavior in the price negotiation phase, such as cost-effectiveness, comparison with other companies, explanatory materials for approval, or discount conditions.

[0053] Furthermore, the same "cautious / reserved" response attribute template can be combined with multiple different scenario behavior templates, such as the "initial hearing phase," "technical verification phase," and "approval phase." This allows for training in situations where a party with the same speaker attributes exhibits different typical behaviors in different negotiation phases.

[0054] Furthermore, the same "price negotiation phase" scene behavior template can be combined with multiple different response attribute templates, such as "cautious / reserved," "aggressive / decisive," and "frequent objectors." This allows for training in how to respond to counterparties with different speaking styles or attitudes regarding typical issues in the same negotiation phase. In this way, by combining N response attribute templates and M scene behavior templates, it is possible to generate up to N x M different role-playing scenarios.

[0055] Furthermore, the response attribute identification information and scene identification information may be received independently via the first and second selection fields as described above, or a combination of the response attribute template and the scene behavior template may be selected together in a single operation. In addition, the prompt generation unit 223 may generate prompts by combining the aforementioned unique information of the sales partner, past negotiation information, or sales materials in addition to the response attribute template and scene behavior template. Even in this case, the fact that the response attribute template, which defines the other party's way of speaking, etc., and the scene behavior template, which defines the other party's behavior in the negotiation scene, are stored separately and selected independently remains unchanged.

[0056] With the above configuration, response attribute templates, which categorize the personal attributes of the other party, such as their way of speaking, and scene behavior templates, which define the typical behavior of the other party during the negotiation phase, are stored separately. By independently selecting and combining these, it is possible to systematically train on the behavior of the other party in typical negotiation scenes that repeatedly occur in sales activities, independently of the other party's personal attributes, without needing to imitate a specific sales partner.

[0057] Although these embodiments have been described above, they are intended to facilitate understanding of the present invention and are not intended to limit its interpretation. The present invention can be modified and improved without departing from its spirit, and equivalents thereof are also included.

[0058] For example, the processing performed by each functional unit of the management server 2 described above may be executed by any of the functional units. Furthermore, different functional units may be added to perform some of the processing performed by each of the functional units described above. Also, the functional units of the management server 2 may be distributed across multiple computers.

[0059] Furthermore, the information stored in each memory unit of the management server 2 may be stored in any of the memory units. That is, the information stored in the multiple memory units mentioned above may be stored in a single memory unit, or a portion of the information stored in one memory unit may be stored in another memory unit.

[0060] <Example 1> In the embodiment described above, an example was shown in which the persona setting unit 212 automatically selects an appropriate persona based on the unique information of the sales partner. However, the user may be allowed to manually select a persona. Specifically, after selecting a sales partner, the system presents the user with a list of persona candidates, allowing the user to select an appropriate persona according to the purpose and situation of the sales negotiation. For example, by allowing the user to select different personas for the same sales partner according to the negotiation phase, such as "a conservative decision-maker during the initial visit," "a cost-conscious purchasing manager during budget negotiations," and "a technology-oriented engineer during the technical verification phase," training that can respond to a wider variety of situations becomes possible. In this case, the specific behavior of the other party according to the negotiation phase, such as the initial visit, budget negotiation, or technical verification, is defined by the scene behavior template described later. The persona defines personal attribute characteristics such as the way of speaking, attitude, or decision-making tendencies of the decision-maker, purchasing manager, or engineer, and does not include specific behavior in the negotiation phase.

[0061] <Modification 2> While the above-described embodiment primarily focused on text-based role-playing, voice-based role-playing can also be achieved by combining speech recognition and speech synthesis technologies. Specifically, the user's voice input is converted into text in real time, and the AI's response is output as voice using speech synthesis technology. Furthermore, by applying different voice characteristics (pitch, speaking speed, accent, etc.) according to the selected persona, a more realistic role-playing experience can be provided. This allows even users who are not comfortable with text input to train effectively, and enables training that more closely resembles actual face-to-face negotiations.

[0062] <Variation 3> While the above-described embodiment assumed one-on-one role-playing, a group role-playing function involving multiple participants can also be added. Specifically, multiple sales representatives can access the system simultaneously, each taking on a different role (e.g., lead salesperson, technical support, manager) and participating in a single sales scenario. The generating AI would then play multiple roles on the client side (e.g., decision-maker, technical staff, finance staff). The system would record each participant's statements and provide individual and overall team evaluations. This would facilitate training in team sales activities and develop the ability to handle complex sales scenarios.

[0063] <Modification 4> In the embodiment described above, an example was shown where the evaluation unit 219 provides evaluation results after the role-playing is completed. However, a function to provide real-time feedback can also be added. Specifically, for each statement made by the user, another generation AI instance performs a real-time analysis and immediately presents the effectiveness of the statement and areas for improvement. This feedback is displayed on a portion of the screen without interrupting the role-playing, or recorded for later review. For example, when a user explains the benefits of a product, feedback such as "Your explanation is not linked to the customer's specific problems" may be provided, or in a price negotiation scenario, positive feedback such as "Presenting alternatives is effective" may be provided. This allows the user to adjust their responses during the role-playing, thereby enhancing the learning effect.

[0064] <Modification 5> The above-described embodiment primarily utilizes proprietary information of sales partners, but it is also possible to add a function that automatically acquires and utilizes industry-specific trends and regulatory information. Specifically, news articles, market research reports, and regulatory change information related to the sales partner's industry are automatically collected from the web and provided to the generating AI. For example, in a role-playing scenario with a customer in the medical device industry, the generating AI would present questions and concerns that reflect the latest medical regulations and insurance reimbursement trends. This enables more realistic and up-to-date role-playing scenarios that reflect the current industry situation, allowing sales representatives to train in addressing industry-specific challenges and concerns.

[0065] <Variation 6> In the embodiment described above, an example was shown in which a persona is identified based on the unique information of the sales partner. However, it is also possible to identify a persona using persona identification information other than the unique information of the sales partner. Specifically, the persona reception unit receives attribute information such as industry category, job title level, decision-making pattern, negotiation phase, or negotiation objective as persona identification information, and the persona setting unit identifies an appropriate persona based on this information. Here, the negotiation phase or negotiation objective is not an attribute of the persona itself, but can be used as scene identification information to identify the scene behavior template described later, and the persona does not include specific behaviors such as questions, concerns, objections, or information to be presented by the other party in these negotiation phases.

[0066] For example, if a user inputs attribute information such as "manufacturing industry," "department head level," "cost-conscious," and "initial proposal," the system will identify a persona corresponding to this combination: "a manufacturing department head level individual who is cost-conscious and is a potential initial proposal recipient." Similarly, if a user specifies the negotiation status as "technical verification phase" and "security-conscious," the system will select a persona as "a technical verification specialist who prioritizes security." Note that negotiation phases such as the initial proposal or technical verification phase can be used not as persona attributes, but as scene identification information to identify the scene behavior templates described later, and the specific behavior of the other party in that negotiation phase is defined by the scene behavior templates.

[0067] Furthermore, users can directly describe the characteristics of a persona. For example, they can input persona characteristics in text format, such as "a cautious person who asks many questions and seeks specific examples" or "a proactive decision-maker who is highly interested in new technologies." The system will then either select the persona that best matches this description from the persona memory or directly reflect the entered description in the prompt.

[0068] This configuration allows for flexible role-playing training tailored to the attributes of potential clients and the circumstances of negotiations, even before specific sales targets have been identified. This is particularly effective in new business development and cross-industry sales activities, enabling sales representatives to efficiently acquire the skills to handle negotiations with various types of customers.

[0069] <Example 7> In the embodiment described above, the user selects a business partner. The user operating user terminal 1 selects a sales partner via the candidate selection UI unit 220 (S501). The information of the selected sales partner is transmitted to the sales partner reception unit 211 (S <Disclosure Items> Furthermore, this disclosure also includes the following configurations. [Item 1] An information processing system that uses generative AI to support sales role-playing, A response attribute template storage unit stores multiple response attribute templates that categorize personal or conversational attribute characteristics, including at least one of the role-playing partner's speaking style, attitude, questioning tendencies, or decision-making tendencies. A scene behavior template storage unit stores, separately from the response attribute templates, multiple scene behavior templates corresponding to the negotiation phase or negotiation objective, each containing at least one of the questions, concerns, objections, requests, disclosure information, non-disclosure information, progress conditions, or termination conditions that the other party should present in the negotiation phase or negotiation objective; A reception unit that receives response attribute identification information that identifies one response attribute template from among the multiple response attribute templates, and scene identification information that identifies one scene action template from among the multiple scene action templates, A prompt generation unit generates a prompt that instructs the generating AI to behave as a role-playing partner, speaking or acting in accordance with the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information, and asking questions, expressing concerns, raising objections, making requests, or disclosing information in accordance with the scene behavior template. An information processing system characterized by comprising the following features. [Item 2] The information processing system according to item 1, wherein the prompt generation unit can combine the same response attribute template with a plurality of different scene action templates, and the same scene action template can combine with a plurality of different response attribute templates. [Item 3] The aforementioned response attribute template does not include names, organization names, URLs, or past sales history that identify a specific sales partner, but is stored as type information applicable to multiple sales partners, as described in Item 1, in the information processing system. [Item 4] The information processing system described in item 1 includes a scene action template which is updated according to the conversation history during the role-playing, and questions, counterarguments, or disclosure information that the generating AI can present according to the scene state. [Item 5] A role-playing reception unit that receives input from users for role-playing, A response generation unit provides the prompt and the input to the generating AI to generate a response to the input, An output unit that outputs the response to the user, The information processing system described in item 1, further comprising the above. [Item 6] The information processing system according to item 5, further comprising an evaluation unit that compares the evaluation items or progress conditions included in the scene action template with the user's input history and evaluates whether the user has performed the necessary confirmation, explanation, or agreement formation during the negotiation phase. [Item 7] The information processing system described in item 1, wherein the reception unit receives the response attribute identification information and the scene identification information via a user interface having a first selection field for selecting the response attribute template and a second selection field for selecting the scene action template. [Item 8] The system further includes a customer information storage unit that stores unique information of the customer in association with the customer, The information processing system described in item 1, wherein the prompt generation unit generates the prompt by combining the response attribute template and the scene action template with the unique information of a specified sales partner. [Item 9] An information processing method that uses generative AI to support sales role-playing, The steps include memorizing multiple response attribute templates that categorize personal or conversational attribute characteristics, including at least one of the role-playing partner's speaking style, attitude, questioning tendencies, or decision-making tendencies, and A step of storing multiple scene behavior templates, corresponding to the negotiation phase or negotiation objective, and including at least one of the questions, concerns, objections, requests, disclosure information, non-disclosure information, progress conditions, or termination conditions that the other party should present in said negotiation phase or negotiation objective, separately from the response attribute templates, A step of receiving response attribute identification information that identifies one response attribute template from among the multiple response attribute templates, and scene identification information that identifies one scene action template from among the multiple scene action templates, A step of generating a prompt that instructs the generating AI to behave as a role-playing partner, speaking or acting in accordance with the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information, and asking questions, expressing concerns, raising objections, making requests, or disclosing information in accordance with the scene behavior template, by combining the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information. An information processing method characterized by a computer executing the following. [Item 10] A program that uses generative AI to support sales role-playing, The steps include memorizing multiple response attribute templates that categorize personal or conversational attribute characteristics, including at least one of the role-playing partner's speaking style, attitude, questioning tendencies, or decision-making tendencies, and A step of storing multiple scene behavior templates, corresponding to the negotiation phase or negotiation objective, and including at least one of the questions, concerns, objections, requests, disclosure information, non-disclosure information, progress conditions, or termination conditions that the other party should present in said negotiation phase or negotiation objective, separately from the response attribute templates, A step of receiving response attribute identification information that identifies one response attribute template from among the multiple response attribute templates, and scene identification information that identifies one scene action template from among the multiple scene action templates, A step of generating a prompt that instructs the generating AI to behave as a role-playing partner, speaking or acting in accordance with the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information, and asking questions, expressing concerns, raising objections, making requests, or disclosing information in accordance with the scene behavior template, by combining the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information. A program that causes a computer to execute something. [Explanation of Symbols]

[0070] 1 User terminal 2 Management Server 3. Sales Management Database 4 Business negotiation management DB 5 Material DB

Claims

1. An information processing system that uses generational AI to support sales role-playing, A response attribute template storage unit stores multiple response attribute templates that categorize personal or conversational attribute characteristics, including at least one of the role-playing partner's speaking style, attitude, questioning tendencies, or decision-making tendencies. A scene behavior template storage unit stores, separately from the response attribute templates, multiple scene behavior templates corresponding to the negotiation phase or negotiation objective, each containing at least one of the questions, concerns, objections, requests, disclosure information, non-disclosure information, progress conditions, or termination conditions that the other party should present in the negotiation phase or negotiation objective; A reception unit that receives response attribute identification information that identifies one response attribute template from among the multiple response attribute templates, and scene identification information that identifies one scene action template from among the multiple scene action templates, A prompt generation unit generates a prompt that instructs the generating AI to behave as a role-playing partner, speaking or acting in accordance with the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information, and asking questions, expressing concerns, raising objections, making requests, or disclosing information in accordance with the scene behavior template. An information processing system characterized by comprising the following features.

2. The information processing system according to claim 1, wherein the prompt generation unit can combine the same response attribute template with a plurality of different scene action templates, and the same scene action template can combine with a plurality of different response attribute templates.

3. The information processing system according to claim 1, wherein the aforementioned response attribute template does not include a name, organization name, URL, or past business negotiation history that identifies a specific business partner, and is stored as type information that can be commonly applied to multiple business partners.

4. The information processing system according to claim 1, wherein the scene action template includes a scene state that is updated according to the conversation history during the role-playing, and questions, counterarguments, or disclosure information that the generating AI can present according to the scene state.

5. A role-playing reception unit that receives input from users for role-playing, A response generation unit provides the prompt and the input to the generating AI to generate a response to the input, An output unit that outputs the response to the user, The information processing system according to claim 1, further comprising:

6. The information processing system according to claim 5, further comprising an evaluation unit that compares evaluation items or progress conditions included in the scene action template with the user's input history and evaluates whether the user has performed the necessary confirmation, explanation, or agreement formation during the negotiation phase.

7. The information processing system according to claim 1, wherein the reception unit receives the response attribute identification information and the scene identification information via a user interface having a first selection field for selecting the response attribute template and a second selection field for selecting the scene action template.

8. The system further includes a customer information storage unit that stores unique information of the customer in association with the customer, The information processing system according to claim 1, wherein the prompt generation unit generates the prompt by combining the response attribute template and the scene action template with the unique information of a specified business partner.

9. An information processing method that uses generative AI to support sales role-playing, The steps include memorizing multiple response attribute templates that categorize personal or conversational attribute characteristics, including at least one of the role-playing partner's speaking style, attitude, questioning tendencies, or decision-making tendencies, and A step of storing multiple scene behavior templates, corresponding to the negotiation phase or negotiation objective, and including at least one of the questions, concerns, objections, requests, disclosure information, non-disclosure information, progress conditions, or termination conditions that the other party should present in said negotiation phase or negotiation objective, separately from the response attribute templates, A step of receiving response attribute identification information that identifies one response attribute template from among the multiple response attribute templates, and scene identification information that identifies one scene action template from among the multiple scene action templates, A step of generating a prompt that instructs the generating AI to behave as a role-playing partner, speaking or acting in accordance with the response attribute template identified by the response attribute identification information and making questions, expressing concerns, raising objections, making requests, or disclosing information in accordance with the scene behavior template, by combining the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information. An information processing method characterized by a computer executing the following.

10. A program that uses generative AI to support sales role-playing, The steps include memorizing multiple response attribute templates that categorize personal or conversational attribute characteristics, including at least one of the role-playing partner's speaking style, attitude, questioning tendencies, or decision-making tendencies, and A step of storing multiple scene behavior templates, corresponding to the negotiation phase or negotiation objective, and including at least one of the questions, concerns, objections, requests, disclosure information, non-disclosure information, progress conditions, or termination conditions that the other party should present in said negotiation phase or negotiation objective, separately from the response attribute templates, A step of receiving response attribute identification information that identifies one response attribute template from among the multiple response attribute templates, and scene identification information that identifies one scene action template from among the multiple scene action templates, A step of generating a prompt that instructs the generating AI to behave as a role-playing partner, speaking or acting in accordance with the response attribute template identified by the response attribute identification information and making questions, expressing concerns, raising objections, making requests, or disclosing information in accordance with the scene behavior template, by combining the response attribute template identified by the response attribute identification information and the scene behavior template identified by the scene identification information. A program that causes a computer to execute something.

Images