Business plan creation system, management analysis method of business plan creation system, and computer program

Abstract

The present invention automatically generates a subsidy application document requested by an application destination while reflecting a learning model, which is obtained by learning and accumulating atypical data to be entered in a subsidy application, to a management analysis in cooperation with each other.　This business plan creation system (1) for creating a business plan in cooperation with a learning model that retains parameters obtained by learning response information for to-be-entered items requested by a subsidy application, wherein the business plan creation system (1) is provided with a management analysis unit (16) that: refers to assumed questions and answers obtained by combining questions pertaining to fundraising and answers to same; acquires answers to questions and financial information from a user terminal (3) by using a learning model (2C) that retains parameters obtained by learning response information for atypical data items requested by the subsidy application; and performs management analysis by using the learning model (2C) on the basis of questions and answers to same.

Description

Business Plan Creation System, Business Analysis Method of Business Plan Creation System, and Computer Program

[0001] The present invention relates to a business plan creation system, a business analysis method of the business plan creation system, and a computer program.

[0002] In business management, it is important to raise funds such as subsidies. For this, it is important to create a business plan for applying for subsidies. Also, investors, partner companies, etc. are always looking for excellent venture companies, etc. Therefore, business operators also need to create a business plan for raising funds.

[0003] However, even if business operators (especially small and medium-sized enterprises, etc.) try to organize the current situation and issues for applying for subsidies, they cannot investigate and organize them as desired.

[0004] Also, they do not know how to formulate revenue forecasts and KPIs (business goals) for applying for subsidies. Furthermore, although they want to raise subsidies, there is no material to convince others of the attractiveness and growth potential of the business.

[0005] Managers of businesses with dozens of employees are so busy dealing with various daily tasks that they do not have time to create a business plan.

[0006] On the other hand, if there is no comprehensive material for applying for subsidies, explaining the business to investors, partner companies, etc., there may be an opportunity to miss raising funds.

[0007] For the company side, they want to quickly summarize a business plan without much effort. In contrast, external parties such as subsidy applicants, investors, or partner companies think that they will provide funds if there is a more comprehensive business plan.

[0008] To solve such problems, for example, Patent Document 1 is disclosed. The skill analysis program of Patent Document 1 provides a skill analysis method for clarifying important skill fields for a company.

[0009] This skill analysis method involves selecting several companies with improved performance from a large number of companies, and calculating the rate of increase in skill levels (skill improvement rate) for each skill area within each of these companies. Then, the average skill improvement rate for each skill area is determined, and this average is set as the standard skill improvement rate for that skill area. The skill improvement rates in each skill area of ​​the company requesting the skill analysis are then compared to the standard skill improvement rate to identify the skill areas that are important to the company but where there is a deficiency, and a plan is developed to address those deficiencies.

[0010] Japanese Patent Publication No. 2003-281326

[0011] However, the skills analysis program described in Patent Document 1 does not include instructions for creating business plans or documents explaining the future prospects of the business. Therefore, there is a problem in that the plan formulated using Patent Document 1 does not lead to a subsidy application.

[0012] Furthermore, conventional subsidy application systems employ a configuration where the server device that stores and analyzes customer management information does not interact with the server device that analyzes the management information of a specific company, and instead provides a unique information processing service.

[0013] Therefore, simply inputting prompts from user devices operated by businesses to some of the generating AI is unlikely to automatically generate the optimal business plan for each business to submit to eligible subsidy recipients. The current challenge is that, in order to develop a truly effective subsidy application service, it is necessary to collaborate with these server devices and perform service processing based on distinctive routines.

[0014] Thus, the information planned for conventional subsidy application systems is primarily biased towards financial information, and no system has been proposed to handle the non-quantitative information that the reviewers of subsidy applications want to know—specifically, information that cannot be expressed in numbers or statistics, such as a company's values, social significance, and future potential.

[0015] Therefore, even if financial information is excellent, it has been pointed out that there is a challenge in automatically generating compelling subsidy applications that reflect non-quantitative information such as corporate philosophy and vision, founding story, social impact, customer testimonials and experiences, future outlook and growth strategy, and organizational culture and values.

[0016] Furthermore, conventional chatbots were limited to automating the handling of inquiries, and lacked a mechanism for utilizing the obtained information for creating business plans or marketing analysis. In other words, evaluating and improving the quality of conversations was subjective, making it difficult to identify conversations that led to conversions.

[0017] This invention was made in view of the above problems, and aims to automatically introduce a storytelling structure that enhances the appeal of application documents, even if the person in charge at a company wishing to apply for a subsidy is not familiar with the application procedures or the details of document writing, and to automatically generate a business plan that conforms to the secretariat's review criteria, which include customer information and management analysis information required by the subsidy recipient, by analyzing the business situation of the business using the system, including unstructured data.

[0018] To solve the above problems, one embodiment of the present invention is a business plan creation system that includes a user terminal that instructs the automatic generation of a subsidy application form to be submitted to a predetermined subsidy application recipient, and a server device that provides a service for creating a business plan for fundraising in cooperation with a learning model that holds parameters learned from response information to the items to be described in the subsidy application form, wherein the server device includes a database that stores the application guidelines for business plans requested by the predetermined subsidy application recipient and customer management information of the service user, and a business analysis unit that refers to a set of hypothetical questions and answers that combine questions and answers regarding fundraising, obtains answers to the questions from the user terminal using the learning model, and performs business analysis using the learning model based on the questions and answers, and a creation unit that instructs the learning model to create a business plan using the customer management information, the application guidelines, the results of the business analysis, and an evaluation score for the answers using a chatbot function with the user terminal, and obtains the business plan created by the learning model based on the parameters.

[0019] Furthermore, another embodiment of the present invention is a business analysis method for a business plan creation system, comprising: a user terminal that instructs the automatic generation of a subsidy application form to be submitted to a predetermined subsidy recipient; and a server device that provides a service for creating a business plan for fundraising in cooperation with a learning model that holds parameters learned from response information to the items to be described in the subsidy application form, wherein the server device includes a database that stores the application guidelines for business plans requested by the predetermined subsidy recipient and customer management information for users of the service, and refers to a set of hypothetical questions and answers that combine questions and answers regarding fundraising, and uses the learning model to obtain answers to the questions from the user terminal and performs a business analysis using the learning model based on the questions and answers; and a creation step instructing the learning model to create a business plan using the customer management information, the application guidelines, the results of the business analysis, and an evaluation score for the answers, using prompts input from the user terminal, and obtaining the business plan created by the learning model based on the parameters.

[0020] Another embodiment of the present invention is a computer program that creates a business plan for fundraising in conjunction with a learning model that holds parameters learned from response information to non-standard data items required by a subsidy application, the program which causes the computer to perform a business analysis step in which it refers to a set of hypothetical questions and answers that combine questions and answers related to fundraising, uses the learning model to obtain answers to the questions from the user terminal, and performs a business analysis using the learning model based on the questions and answers, and instructs the learning model to create a business plan using customer management information, the application guidelines, the results of the business analysis, and an evaluation score for the answers, using prompts input from the user terminal, and obtains the business plan created by the learning model based on the parameters.

[0021] As described above, the present invention provides the effect of enabling the creation of a system that automatically generates subsidy application documents tailored to the application form of the recipient, and automatically submits them to selected recipients, without requiring company representatives seeking subsidies to be familiar with the application procedures or the details of filling out application documents. This system actively incorporates a storytelling structure to enhance the appeal of the application documents, thereby improving persuasiveness and empathy, and increasing the acceptance rate.

[0022] The drawings illustrate specific embodiments of the present invention, including not only essential components of the invention but also selective and preferred embodiments. This is a block diagram illustrating the functions of a business plan creation system according to one embodiment of the present invention. This is a configuration diagram showing the server configuration of a business plan creation system according to one embodiment of the present invention. This is a configuration diagram showing an example of the hardware configuration of the main server of a business plan creation system according to one embodiment of the present invention. This is a configuration diagram showing an example of the hardware configuration of the server related to customer management of a business plan creation system according to one embodiment of the present invention. This is a configuration diagram showing an example of the hardware configuration of another server related to customer management of a business plan creation system according to one embodiment of the present invention. This is a configuration diagram showing an example of the hardware configuration of the server related to the AI ​​chatbot of a business plan creation system according to one embodiment of the present invention. This is an explanatory diagram illustrating the data configuration according to one embodiment of the present invention. This is a flowchart illustrating the processing selection according to one embodiment of the present invention. This is an explanatory diagram illustrating the TOP screen according to one embodiment of the present invention. This is a flowchart illustrating the operation of the AI ​​chatbot processing according to one embodiment of the present invention. This is an explanatory diagram illustrating an example of the screen of the AI ​​chatbot processing according to one embodiment of the present invention. This is an explanatory diagram illustrating another a flowchart illustrating the business analysis processing according to one embodiment of the present invention. This is an explanatory diagram illustrating an example of the screen of the business analysis processing according to one embodiment of the present invention. This is an explanatory diagram illustrating another example screen for the business analysis process according to one embodiment of the present invention. This is an explanatory diagram illustrating another example screen for the business analysis process according to one embodiment of the present invention. This is an explanatory diagram illustrating another example screen for the business analysis process according to one embodiment of the present invention. This is a flowchart illustrating the business plan creation process according to one embodiment of the present invention. This is an explanatory diagram illustrating an example screen for the output of a business plan document according to one embodiment of the present invention. This is an explanatory diagram illustrating an example output of the business plan creation process according to one embodiment of the present invention. This is a flowchart illustrating the subsidy information delivery process according to one embodiment of the present invention. This is an explanatory diagram illustrating an example screen for the subsidy information delivery process according to one embodiment of the present invention.This is an explanatory diagram illustrating another example screen for the subsidy information delivery process according to one embodiment of the present invention. This is a flowchart illustrating the subsidy application review process according to one embodiment of the present invention. This is an explanatory diagram illustrating an example screen for the business plan review process according to one embodiment of the present invention. This is an explanatory diagram illustrating another example screen for the business plan review process according to one embodiment of the present invention. This is a diagram showing an example of the operation screen of the business plan creation support system according to one embodiment of the present invention. This is a configuration diagram showing an example of the hardware configuration of the main server of the business plan creation support system according to one embodiment of the present invention. This is a flowchart illustrating the operation related to AI chatbot processing according to one embodiment of the present invention. This is a diagram showing an example of the user interface screen displayed on the display unit of the user terminal shown in Figure 1 Figure 1 shows an example of the management philosophy information generation screen displayed on the user terminal's display unit. Figure 1 shows an example of the business plan creation instruction screen for subsidy applications displayed on the user terminal's display unit.

[0023] The embodiments will be described in detail below with reference to the attached drawings. In this example, publicly known technologies will not be explained. Furthermore, the structures and methods described are illustrative examples for realizing the technical idea of ​​the invention, and the technical idea of ​​the present invention is not limited to those described below. The technical idea of ​​the present invention can be modified in various ways within the scope of the claims. In particular, it should be noted that the drawings are schematic and may differ from reality. The embodiments represent the most preferred form of the invention, and the invention is not limited thereto.

[0024] Corporate fundraising methods include subsidies, grants, loans, and capital investments. Below, we will use subsidy applications as an example of fundraising. A business plan for subsidy applications clearly outlines, for example, "what the company should be like in the future," along with management strategies and action plans. It presents a concrete vision and strategy for the company's future to people both inside and outside the company, such as subsidy providers, employees, business partners, investors, and financial institutions.

[0025] Creating a business plan for a subsidy application requires a thorough examination of market trends, the company's position, the development of a business plan, and the necessary funding. This allows for a clear understanding of the current situation surrounding the company and makes it easier to identify its strengths and weaknesses in the market. Therefore, this information is crucial for developing a business strategy.

[0026] Having a solid business plan for grant applications makes it easier to create a more accurate financial plan.

[0027] [First Embodiment] First, the overall configuration of this embodiment will be described with reference to Figures 1 and 2. Figure 1 is a block diagram showing the functions of the business plan creation system according to this embodiment, and Figure 2 is a configuration diagram showing the server configuration of the business plan creation system according to this embodiment. Here, the business plan is information consisting of items such as company overview, background, management challenges, market analysis, competitor analysis, analysis methods, and problem identification.

[0028] As shown in Figure 1, the business plan creation system 1 consists of a learning model 2C, a user terminal 3, a CB information DB (hereinafter referred to as CB information DB) 11A, a CB management unit 11B, a customer management DB 12, a customer management unit 13, an application document DB 14A, an application management unit 14B, a business plan creation / review unit 15, a management analysis unit 16, an application management DB 17, an output conversion unit 18, an application procedure unit 19, an operation unit 20, and the like.

[0029] 2 is the generation AI core unit, which consists of a verification unit 2A and a machine learning unit 2B that includes the learning model 2C. The machine learning unit 2B optimizes the parameters of the learning model 2C using training data. Here, the training data includes the application guidelines, evaluation criteria, company information, evaluation comments, and conversations obtained by users who have already used this system through question-and-answer sessions with the chatbot unit 11C when creating a subsidy application, as required by the items to be described in the subsidy application form. However, it may also include other information. Furthermore, the contents of accepted and rejected subsidy applications may also be used as training data.

[0030] Furthermore, "approved application forms" refer to past grant applications that have been approved (in PDF or Word® format), allowing the learning model 2C to learn successful application patterns.

[0031] Furthermore, the reason why rejected applications are used to train the machine learning unit 2B is so that the learning model 2C can identify failure patterns. For example, since the intended use of each grant is predetermined, the learning model 2C can detect whether there are any omissions in the application documents that the reviewers focus on during the review process.

[0032] Furthermore, the reason why the machine learning unit 2B is trained on the application guidelines and review process is so that the learning model 2C can identify evaluation items and scoring factors for each grant. This allows for the optimization of the output of the generated application documents (reflecting the reviewer's perspective).

[0033] Furthermore, the reason why the chatbot unit 11C is trained by the machine learning unit 2B with the company philosophy, business plan, financial information (B / S), etc., that it receives from the user is so that the learning model 2C can make decisions on extracting and integrating non-quantitative information.

[0034] Furthermore, the reason why the machine learning unit 2B is trained on the review comments is so that the learning model 2C can incorporate any feedback from the reviewers into the generation of the next application documents.

[0035] Furthermore, the reason why the machine learning unit 2B learns the template structure is to determine whether the learning model 2C conforms to the format set by the application recipient.

[0036] Furthermore, the learning model 2C holds the parameters learned by the machine learning unit 2B and performs inference processing on the input information. Specifically, this inference processing includes inference processing related to natural language generation, image generation, and structured data generation.

[0037] The verification unit 2A evaluates the output of the learning model 2C for accuracy, validity, safety, etc. The verification unit 2A may also provide feedback to the machine learning unit 2B based on the evaluation results. Furthermore, this system includes a function to attach documents to the application form specified by the application recipient. These documents, while not directly noticed by the application recipient's examiner, serve as a way for the company to showcase itself—in other words, a presentation (a good example is a response simulating an interview with an examiner).

[0038] In this embodiment, the learning model 2C is GPT®, Gemini®, etc., which generate answers to questions (prompts, attached documents). The functions of the learning model 2C will be described later.

[0039] Thus, this system is characterized by the fact that, in order to facilitate the routine of creating subsidy applications, the training data is composed of diverse information sources, is structured and pre-processed rather than simply being a collection of documents, and plays a role in directly contributing to improved output quality. Structured and pre-processed means that, for example, when machine learning is performed on company name, management philosophy, history, and founding year, a hierarchy such as company name → history → founding year → management philosophy (this hierarchy is just one example) is set up, and the information associated with it is linked and stored.

[0040] Note that in the following creation of the subsidy application form, on the premise that the above learning process has been performed, the details of the subsidy application process will be described mainly based on the learning model 2C.

[0041] The user terminal 3 is a device operated by the user who applies for the subsidy. This user terminal 3 is a mobile terminal such as a smartphone or a PC, and in addition to a CPU and a communication device (not shown), it includes an input unit 31 for operation input, a display unit 32 for displaying characters, images, etc. Of course, the input unit 31 and the display unit 32 may be an integrated touch panel.

[0042] The CB information DB 11A stores, in tabular form, an assumed Q&A set combining questions regarding fundraising (including subsidy applications) and their answers. The CB management unit 11B has an AI chatbot function using the learning model 2C. This CB management unit 11B refers to the CB information DB 11A, receives questions regarding fundraising from the user via the user terminal 3 and answers them, scores the evaluation of the answers, and stores the obtained scores in the CB information DB 11A.

[0043] The customer management DB 12 stores customer management information obtained by CRM (Customer Relationship Management) and MA (Marketing Automation) to be described later. The customer management information will be described in detail in the explanation of FIG. 7. The customer management unit 13 collects the above-mentioned customer management information and stores it in the customer management DB 12, and provides the customer management information to the business plan creation / review unit 15 when creating or reviewing a business plan.

[0044] The application document DB 14A stores the solicitation requirements for the business plan required for subsidy applications and other applications in fundraising. The application management unit 14B can newly add to the application document DB 14A by the user's operation via the user terminal 3, and as will be described later, manages the data of the business plan created for the application in the application management DB 17. The application management unit 14B further receives an instruction from the user via the user terminal 3 and executes the application procedure by the application procedure unit 19.

[0045] The business plan creation / review section 15 creates a business plan based on the assumed Q&A set and scores stored in the CB information DB 11A, the customer management information stored in the customer management DB 12, and the business analysis results obtained by the business analysis section 16 using the learning model 2C, and stores the business plan in the application management DB 17.

[0046] The business analysis section 16 generates questions for filling in each item of the solicitation requirements stored in the application document DB 14A with appropriate information using the learning model 2C, and stores the user's answers to the questions in the application document DB 14A linked to the corresponding questions.

[0047] The application management DB 17 stores the data of the business plan created by the business plan creation / review section 15 linked to information on the funding source and the user. When the output conversion section 18 receives an output request from the user via the user terminal 3, it converts the data of the business plan stored in the application management DB 17 into, for example, the WORD (registered trademark) format and outputs it to the user terminal 3.

[0048] The application procedure section 19 creates and transmits application documents to the terminal of the destination for applying the business plan based on the data of the business plan stored in the application management DB 17. The operation section 20 performs operations such as editing the assumed Q&A set stored in the CB information DB 11A and checking the score for the CB management section 11B. The operation section 20 receives operations from the user via the user terminal 3 and performs operations related to CRM and MA for the customer management section 13. The operation section 20 performs operations such as registering new solicitation requirements for the application management section 14B.

[0049] As shown in FIG. 2, the business plan creation system 1 is configured such that the user terminal 3, the server 5, the CRM server 6, the MA server 7, the AICB server 8, and the learning model 2C are connected to the network 4. The user terminal 3 is a representative example of the device operated by the person in charge of each company.

[0050] The CRM server 6 has a configuration that includes the customer management unit 13 and customer management DB 12 as shown in Figure 1, and similarly, the MA server 7 has a configuration that includes the customer management unit 13 and customer management DB 12 as shown in Figure 1. The server 5 has a configuration that includes the application document DB 14A, application management unit 14B, business plan creation / review unit 15, management analysis unit 16, application management DB 17, output conversion unit 18, application procedure unit 19, operation unit 20, etc. as shown in Figure 1. The AICB server (hereinafter referred to as the AICB server) 8 has a configuration that includes the CB information DB 11A, CB management unit 11B, etc.

[0051] Next, we will explain the individual server configurations with reference to Figures 3 to 6. Figures 3, 4, 5, and 6 are configuration diagrams showing examples of the hardware configurations of Server 5, CRM Server 6, MA Server 7, and AICB Server 8 according to this embodiment, respectively.

[0052] Server 5 is the main device of the business plan creation system 1 and consists of a CPU 51, ROM 52, RAM 53, display 54, keyboard 55, storage 56, etc. The CPU 51 controls the business plan creation support of this embodiment according to the program stored in ROM 52 (the program responsible for the operations shown in Figures 8 to 13).

[0053] ROM 52 stores the above program, and RAM 53 functions as a work area when the CPU 51 is operating. Display 54 is a device that forms the necessary screen as needed during the operation of the server 5, and keyboard 55 is a device used by the operator of the server 5 to input data. Storage 56 consists of the application document DB 14A, application management DB 17, etc., as described above, and data input / output is controlled by the CPU 51.

[0054] The CRM server 6 is a device responsible for CRM processing and consists of a CPU 61, ROM 62, RAM 63, storage 64, and the like. The CPU 61 controls the operation of the entire CRM server 6 in conjunction with the server 5 according to the program stored in the ROM 62.

[0055] ROM 62 stores the above program, and RAM 63 functions as a work area during CPU 61 operation. Storage 64 is composed of the CRM management DB 12A and the like, and data input / output is controlled by CPU 61.

[0056] The MA server 7 is a device responsible for MA processing and consists of a CPU 71, ROM 72, RAM 73, storage 74, etc. The CPU 71 controls the operation of the entire MA server 7 in conjunction with the server 5 according to the program stored in the ROM 72.

[0057] ROM 72 stores the above program, and RAM 73 functions as a work area during CPU 71 operation. Storage 74 is composed of the MA management DB 12B and the like, and data input / output is controlled by CPU 71.

[0058] The AICB server 8 is a device responsible for AI chatbot processing and consists of a CPU 81, ROM 82, RAM 83, storage 84, and the like. The CPU 81 controls the operation of the entire AICB server 8 in conjunction with the server 5 according to the program stored in the ROM 82.

[0059] ROM 82 stores the above program, and RAM 83 functions as a work area during CPU 81 operation. Storage 84 is composed of the CB information DB 11A and the like, and data input / output is controlled by CPU 81.

[0060] Next, the configuration of each database will be explained with reference to Figure 7. In Figure 7, (A) shows the data configuration of CB Information DB11A, (B) shows CRM Management DB12A, (C) shows MA Management DB12B, (D) shows Application Document DB14A, and (E) shows Application Management DB17.

[0061] As shown in Figure 7(A), the CB information DB11A consists of a business ID that identifies a pre-registered business, a business name, a set of anticipated questions and answers, a score associated with each set of questions and answers, and the date and time of the last update.

[0062] As shown in Figure 7(B), the CRM management DB12A consists of the business ID, business name, and customer management information, sales promotion information, marketing information, contract-related information, and analytical information for each business obtained through CRM processing.

[0063] As shown in Figure 7(C), the MA management DB12B consists of the business ID, business name, business-specific customer management information, behavioral history information, segment information, campaign information obtained through MA processing, score information, and analysis information obtained through MA processing evaluation.

[0064] In the case of a subsidy application, the application document DB14A consists of an ID that identifies the organization applying for the subsidy, the subsidy name indicating the title of the subsidy application, information on the application guidelines indicating the specific items required for the subsidy application, and input information required for each item of the application guidelines, as shown in Figure 7(D). This input information indicates the items and includes information obtained from AI chatbots, CRM, MA, homepage, financial statements, company profiles, and APIs for linking with external devices. For each item of the application guidelines that is referenced in this input information, the input information and the item are stored and managed in association.

[0065] As shown in Figure 7(E), the application management DB17 consists of the business ID, business name, business plan stored in a predetermined data format, creation date and time, etc.

[0066] Here, we will explain the processing overview of the learning model 2C. In this embodiment, for (1) AI chatbot analysis, (2) business analysis, and (3) creation of a business plan, when a combination of prompts and necessary materials is input in the form of a question, the learning model 2C generates an answer in the following manner.

[0067] (1) When the AI ​​chatbot answers a user's question using the anticipated Q&A set, an evaluation is obtained from the user (meaning user terminal 3). This set of question, answer, and evaluation is scored. The learning model 2C is used for this scoring.

[0068] (2) The content to be analyzed differs depending on the application guidelines for business analysis. Therefore, learning model 2C generates a set of questions for business analysis based on the application guidelines. The set of questions and answers from the user for each question in that set of questions is obtained for the creation of the business plan.

[0069] (3) To briefly summarize the processing of the business plan creation learning model 2C, the set of questions, answers, and scores (bad rate) from the AI ​​chatbot can measure the level of understanding required to create a business plan. The results of the business analysis, which consist of a set of questions and answers related to management, provide information that grasps the actual state of management, including numerical data. The learning model 2C then maps the components in accordance with the application guidelines based on the results from the AI ​​chatbot and the business analysis, generates a document based on the logical structure (NLG) and adapts the style (Style Transfer) according to the submission format, and outputs it after formatting the output format (Markdown, Word®). In addition, the judges' scoring criteria will also be taken into consideration when creating this business plan.

[0070] Next, the operation will be explained with reference to Figures 8 to 27. First, the user selects a process from the TOP screen via the user terminal 3. Figure 8 is a flowchart illustrating the process selection according to this embodiment, and Figure 9 is an explanatory diagram illustrating the TOP screen according to this embodiment.

[0071] First, when the user requests the TOP screen via the user terminal 3 (step S101), the server 5 prepares the display data for the TOP screen 91 shown in Figure 9 and sends (provides) it to the user terminal 3 (step S201). On the user terminal 3, the TOP screen 91 received from the server 5 is displayed on the display unit 32 (step S102). As shown in Figure 9, the TOP screen 91 displays four items that can be selected by the input unit 31: "Business plan creation support," "AI chatbot," "Customer management / sales promotion system," and "Subsidy information delivery."

[0072] When an item to be operated on is selected from the menu on the TOP screen 91 (step S103), the selected item is sent to the server 5 (step S104). The server 5 receives the selection instruction and determines which item has been selected (step S202). If the logout button 91A is pressed, the logout process is executed.

[0073] If the system determines that the "AI chatbot" has been selected, the process moves to the AI ​​chatbot processing shown in Figure 11. If the system determines that the "customer management and sales promotion system" has been selected, the process moves to the CRM / MA processing, although this is not shown in the diagram.

[0074] If the system determines that "Business plan creation support" has been selected, the process moves to the business plan creation support process shown in Figure 14. Furthermore, if the system determines that "Subsidy information delivery" has been selected, the process moves to the business plan creation support process shown in Figure 22.

[0075] [AI Chatbot Processing] Next, the AI ​​chatbot processing will be explained with reference to Figures 10 to 13. Figure 10 is a flowchart illustrating the operation of the AI ​​chatbot processing according to this embodiment, and Figures 11 to 13 are explanatory diagrams illustrating example screens of the AI ​​chatbot processing according to this embodiment.

[0076] In the AI ​​chatbot processing of step S312, as shown in Figure 10, the server 5 and the AICB server 8 communicate via API (steps S211 and S311). In response to the user's question input, the AICB server 8 generates an answer using the learning model 2C, which is then displayed on the user terminal 3. Furthermore, the user inputs an evaluation, and scoring is performed (step S111). The above processes are repeated unless there is a command to return to the TOP screen (no route in step S112).

[0077] Then, when the user input is received and the button 93A returns to the TOP screen 93 is pressed, the processing of the user terminal 3 returns to step S102 described above (YES route of step S112). Then, the cooperation between server 5 and AICB server 8 is terminated and termination processing is executed (steps S212 and S313). At this time, server 5 returns to the processing of step S201, and AICB server 8 enters a waiting state until the next cooperation.

[0078] Regarding the AI ​​chatbot, the history of questions from each user is registered in the CB information DB 11A, linked to the access date and time. As shown in Figure 11, the operator of Server 5 can display a list of the history shown in the CB information DB 11A on the display 54.

[0079] Furthermore, the operator of server 5 associates the answers with each question and registers them in the CB information DB 11A along with the data of the last updated date and time. The operator of server 5 can display the question and answer sets on the display 54, as shown in screen 94 in Figure 12, and can also modify them. The back button 93A remains displayed.

[0080] Figure 13 shows screen 95, which displays a list of questions and answers previously entered by the user through the AI ​​chatbot, and also handles the learning model 2C. The back button 93A remains visible. Although not shown in the figure, the display screen includes the date and time of the question, the content of the answer, the score evaluation (success rate), and the user's feedback, allowing the user to ask further questions or request revisions based on each history. Furthermore, the output results of the AI ​​chat are automatically linked as an important source of information used when creating a business plan. This allows the user to visually grasp the history of changes in their intentions and management policies, leading to the creation of a more persuasive plan.

[0081] [Business Analysis Processing] Next, the business analysis processing will be explained with reference to Figures 14 to 18. Figure 14 is a flowchart illustrating the business analysis processing according to this embodiment, and Figures 15 to 18 are explanatory diagrams illustrating example screens of the business analysis processing according to this embodiment.

[0082] In the business plan creation process, as shown in Figure 14, the user first performs a business analysis. The server 5 provides an initial screen to the user terminal 3 (step S221). On the user terminal 3, as shown in Figure 15, the initial screen 96 is displayed on the display unit 32, and user operations are accepted (step S121). The initial screen 96 includes a button 96A to return to the TOP screen, an AI generation button 96B to instruct the creation of a business plan using the learning model 2C, a progress rate display area 96C showing the progress rate of the business plan, and a subsidy selection area 96D for selection or upload by the user.

[0083] Here, the progress rate display area 96C is automatically calculated based on the input status of each item required in the application guidelines for the grant being applied for (e.g., company overview, challenges, solutions, market analysis, etc.). The completion status of each item is evaluated by both AI judgment (syntactic consistency, descriptive completeness, etc.) and explicit completion actions by the user, and the progress rate is updated in real time. This allows the user to immediately identify any missing information and contribute to the completion of a high-quality business plan without any omissions.

[0084] In the subsidy selection area 96D, as shown in Figure 15, users can select the application guidelines from a dropdown menu. For example, the options "Manufacturing Subsidy" which shows already registered application guidelines, and "Upload Application Guidelines" which allows uploading unregistered application guidelines are displayed.

[0085] When a user selects the application guidelines to be submitted via the user terminal 3 (including uploading them) (step S122), the server 5 creates the application screen 97 shown in Figure 16 according to the selection instructions (step S222), and the display data of that application screen is provided to the user terminal 3 (step S223).

[0086] On the user terminal 3, the application screen 97 received from the server 5 is displayed, and user operations are accepted (step S123). On the application screen 97, an application document creation button 97A and the current business plan are placed in the business plan display area 97B. Then, user operations are sent to the server 5 for buttons etc. on the screen (step S124), and the server 5 makes a processing determination according to the operation input (step S224).

[0087] If the result of the determination is the operation of the AI ​​generation button 96B, the process proceeds to step S225; if the result is the operation of the application document creation button 97A, the process proceeds to the process shown in Figure 19; or if the result is the operation of the button 96A to return to the TOP screen, the process proceeds to step S201 as described above.

[0088] In step S225, as shown in Figure 17, a business analysis screen 98 with a question and answer area 98A is created and provided to the user terminal 3. The user terminal 3 accepts user operations on this business analysis screen 98. The user terminal 3 displays the user's operation input for the question and answer area 98A on the business analysis screen 98, and its contents are shared with the server 5 (step S125). The process in step S125 is repeated until an output instruction is received from the user (no route of step S126). User operation input includes screen scrolling as well as answers (text) to questions necessary for business analysis. In this way, the data necessary for business analysis is gathered.

[0089] When the user operates the analysis result output button 99A (output instruction) on the screen 99 shown in Figure 18 (YES route in step S126), a business analysis based on the answers to the questions in the question and answer area 98A is performed using the learning model 2C. The analysis results are stored in the application document DB 14A. Furthermore, the results are provided to the user terminal 3 for confirmation (step S226). In this way, the server 5 proceeds to step S201. On the user terminal 3, the business analysis results are displayed on the display unit 32 (step S127), and the process proceeds to step S102.

[0090] [Business Plan Creation Process] Next, the business plan creation process will be explained with reference to Figures 19 to 21. Figure 19 is a flowchart illustrating the business plan creation process according to this embodiment, Figure 20 is an explanatory diagram illustrating an example screen for outputting a business plan document according to this embodiment, and Figure 21 is an explanatory diagram illustrating an example output of the business plan creation process according to this embodiment.

[0091] When the AI ​​generation button 96B is pressed on the application screen 97 shown in Figure 16, the process shown in Figure 19 is executed. On server 5, the business analysis results are read (step S231), CRM information is read in cooperation with CRM server 6 (steps S232 and S233), and MA information is read in cooperation with MA server 7 (steps S234 and S235).

[0092] Once all the above data is collected, the application documents, namely the business plan, are created using the learning model 2C (step S236). The business plan created in this way is stored in the application management DB 17. Furthermore, it is provided to the user terminal 3 so that it can be viewed on the user terminal 3 (step S237). On the server 5, the process then moves to step S201. On the user terminal 3, as shown in Figure 20, the AI-generated business plan is displayed on the display unit 32 in the business plan display area 97B (step S131), and the process moves to step S102. Although not shown in the flowchart, an output button 97E is located there.

[0093] When the output button 97E is pressed by the user, a file 100 in Word® format, for example, is output to the user terminal 3, as shown in Figure 21. The user terminal 3 can then download the file 100 and display it on the screen using the Word® application.

[0094] Furthermore, when creating a business plan using the learning model 2C (steps S236 and S237), as will be described later, the table shown in the essential review screen 102 of the business plan is created and registered in the application management DB 17, as shown in Figures 26 and 27.

[0095] [Subsidy Information Delivery Process] Next, the subsidy information delivery process will be explained using Figures 22 to 24. Figure 22 is a flowchart illustrating the subsidy information delivery process according to this embodiment, Figure 23 is an explanatory diagram illustrating an example screen of the subsidy information delivery process according to this embodiment, and Figure 24 is an explanatory diagram illustrating another example screen of the subsidy information delivery process according to this embodiment.

[0096] In the subsidy information delivery process, as shown in Figures 22 and 23, first, a search screen 101 for searching for the names of available subsidies is created on the server 5 and provided to the user terminal 3 (step S241). The user terminal 3 displays the search screen 101 and accepts user input (step S141). For search operations, as shown in Figure 23, there are two methods: entering search keywords in the search area 101A through user operation on the search screen 101, and selecting the desired prefecture from a pull-down menu.

[0097] On server 5, for example, if a region is selected by the user (step S142), a list 101C is created as a search result for that region, listing the names of subsidies published in that region. As a search result, list 101C is provided to user terminal 3 (step S242). On user terminal 3, as shown in Figure 24, list 101C is displayed (step S143), and the user can select the names of subsidies they wish to apply for, for example, by checking checkboxes, with no limit on the number of selections. Furthermore, the application button 101B can be used at any time after the check operation (step S144).

[0098] When the user terminal 3 receives an instruction to operate the application button 101B (YES route in step S145), the name of the subsidy checked in the checkbox is transmitted to the server 5 (step S146). The server 5 then executes an application procedure to send the already prepared business plan in a predetermined file format to the institution of the transmitted subsidy name (step S243).

[0099] When the application is completed on server 5, a notification of application completion is output to user terminal 3 (step S244). User terminal 3 receives the application completion notification (step S147). The application completion notification may be displayed directly on the screen, or it may be sent as an email to the registered email address.

[0100] [Subsidy Application Review Process] Next, the subsidy application review process will be explained with reference to Figures 25 to 27. Figure 25 is a flowchart illustrating the subsidy application review process according to this embodiment, and Figures 26 and 27 are explanatory diagrams illustrating example screens for the business plan review process according to this embodiment.

[0101] In this embodiment, it is possible to review and easily modify an already prepared business plan for subsidy applications. First, the header table data stored in a tabular format as shown in Figure 26 is read from the application management DB 17 and provided to the user terminal 3 (steps S251 and S252). As shown in Figure 26, the header table is configured such that each item of the business plan is associated with the input information that serves as the basis for creating that business plan.

[0102] The input information specifically includes website information (HP), CRM / MA information, AI chatbot information, financial statements, company profiles, and APIs for linking with external functions such as CRM servers and MA servers. The items include company overview, background, management challenges, market analysis, competitor analysis, analysis methods, and problem identification. For example, sub-items within the company overview might include basic information and management philosophy.

[0103] Then, the user terminal 3 receives the header table provided by the server 5 and displays it on the screen as shown in Figures 26 and 27 (step S151). When the user selects an item they want to review, that item is provided by the server 5 and displayed in detail (steps S152 and S253).

[0104] Then, when a user instructs a correction (YES route in step S153), the correction is reflected (steps S154 and S254). Until a correction instruction is received, the process returns to step S152 (NO route in step S153). For this review and correction, processing is executed according to the user's operation via the user terminal 3. If it is CB information, the correction is made in cooperation with the AICB server 8; if it is CRM information, the correction is made in cooperation with the CRM server 6; and if it is MA information, the correction is made in cooperation with the MA server 7. Then, it is possible to recreate and reconfirm the business plan as described above. Furthermore, by being able to review and correct in this way, it is possible to refine the plan.

[0105] Here, I will provide some additional explanation regarding CRM and MA. CRM, which stands for "Customer Relationship Management" in Japanese, is a marketing method for managing and analyzing customer information and building and maintaining good relationships with customers. The purpose of CRM is to maximize LTV (Customer Lifetime Value) by making appropriate approaches to each individual customer, thereby improving the company's profits.

[0106] In a narrower sense, CRM refers to customer management systems and tools. By utilizing CRM, customer information can be analyzed in real time, and services and products can be delivered to each individual at the best possible time, resulting in effective marketing. As a result, it is expected to lead to improved customer satisfaction, increased repeat purchases, enhanced brand reputation, and increased sales.

[0107] Furthermore, MA stands for "a system that automates tasks related to marketing initiatives to streamline marketing activities." In business, while many tools (MA tools) have been developed to realize these marketing activities, it has become common to refer to the tools themselves as MA. With the spread of the internet and smartphones, lead acquisition channels have become more complex, and there is a need to nurture leads according to their needs. Against this backdrop, marketing initiatives have also diversified, and there are limits to what can be managed and implemented manually. This is where MA (Marketing Automation) comes in.

[0108] Furthermore, the following supplementary explanation is provided regarding the business plan created by Business Plan Creation System 1. The items include company overview, background, management challenges, market analysis, competitor analysis, analytical methods, and identification of challenges. The sub-items for company overview include basic information, management philosophy, management policies, added value, types of businesses and services of trading partners, number of trading partners, sales overview, and products and services.

[0109] The background section includes the company's history of operations, achievements, setbacks, crises faced, and the circumstances leading to the subsidy application. The management challenges section includes the company's current financial situation, sales, selling and administrative expenses, asset status, and methods for revising the business plan. The market analysis section includes market characteristics, current market share, current trading area, market size, number of potential customers, market growth rate, personas, and customer needs.

[0110] The competitive analysis sub-items include information on other companies in the market area, points of contention for market share, and points of differentiation from competitors. The analytical method sub-items include SWOT analysis, VRIO analysis, 5force analysis, STP analysis, 3C analysis, and 4P analysis. The issue sub-item includes conducting analysis from the perspective of management issues and customer competition.

[0111] Furthermore, while the creation of an estimate is required for subsidy applications, the business plan creation system 1 can automatically create one. Specifically, the basis for the AI's ability to automatically create an estimate from the business plan is that it references CRM / MA as reference data for the estimate, and the customer's requests are referenced from the AI ​​chatbot history and the automatically created business plan for subsidy applications. Based on the above, server 5 automatically creates the requirements for obtaining an estimate from the business plan in cooperation with the AI ​​model, learning model 2C.

[0112] Furthermore, each fundraising effort requires the submission of project results to investors, funders, and grant secretariats. Various public programs (loans and grants) have specific formats for reporting performance. Therefore, by applying the ability to summarize the application guidelines in the business plan creation system 1 and include headings and details in the business plan, performance reports are also output from the CRM and chatbot databases. The system is then configured to allow users to select the appropriate public format for submitting project results to investors, funders, and grant secretariats.

[0113] When obtaining quotations, it may be possible to refer to the Legal Affairs Bureau database for basic company information such as the corporate number, company name, and representative of the company to which the requirements specification is submitted, while also requiring prospective participating companies to submit a complete commercial history certificate to verify their actual status.

[0114] [Effects of the First Embodiment] As described above, according to this embodiment, even those who are not good at understanding the application guidelines for subsidies or the systems for fundraising through loans or stocks can use AI to provide input support and business analysis, and automatically generate a business plan that conforms to the secretariat's review criteria.

[0115] [Second Embodiment] Figure 28 shows an example of the operation screen of a business plan creation support system according to one embodiment of the present invention. The difference between the screen configuration shown in Figure 28 and the screen configuration shown in Figure 9 is that the management analysis system is linked to it.

[0116] [System Configuration of This Embodiment] Figure 29 is a configuration diagram showing an example of the hardware configuration of the main server of a business plan creation support system according to one embodiment of the present invention.

[0117] In Figure 29, 11C is the chatbot unit, which, in accordance with instructions from the CB management unit 11B, displays a chat screen on the display unit 32 in conjunction with the user's operation instructions on the display screen shown in Figure 28, and provides answers to the user's questions and advice regarding subsidy applications. The information received in response is stored as customer information in the CB information DB 11A.

[0118] Here, the CRM data structure includes columns designed for reporting subsidy performance, allowing for the identification of whether each transaction record is related to a specific subsidy. This feature is intended to allow the system to be used independently, even if the customer does not have their own CRM.

[0119] Server 5 automatically collects publicly available subsidy information by crawling or scraping the web. This process is designed to mimic how a human would search for something like "Minato Ward Small and Medium-Sized Enterprise Subsidy" on a search engine.

[0120] From the text information in the customer information acquired by the CRM server 6, the learning model 2C extracts the following main data (1) to (4).

[0121] (1) Amount of subsidy (2) Eligible expenses (3) Application deadline (4) Subsidy rate (e.g., 2 / 3, 1 / 2) Here, the data (1) to (4) extracted by the learning model 2C are reflected in the data table in the storage 64 within the system.

[0122] In this state, when a user (person in charge) on user terminal 3 selects a specific subsidy from the subsidy list and presses the application button to create a business plan, server 5 determines the status of the application button press and the process starts. Triggered by this application action, CRM server 6 sets a flag in the related transaction record linked to the customer management information it manages, indicating which subsidy it is associated with. If the user does not apply for a subsidy, CRM server 6 continues to store the related transaction record in the database without setting the flag.

[0123] On the other hand, when a user presses the application button displayed on the display unit 32 of the user terminal 3, the corresponding transaction record with the flag set is imported for subsequent processing.

[0124] This automates the process of collecting and updating subsidy information, preventing applications from being missed.

[0125] Furthermore, by linking with the CRM function provided by CRM Server 6, consistency between application content and performance reports can be ensured.

[0126] Furthermore, even if the user terminal 3 does not have a CRM system, the user terminal 3 can use the CRM server 6 as its own CRM function resource.

[0127] Furthermore, Server 5 can improve the traceability of subsidy utilization by linking application history and transaction history as a system.

[0128] [Example of business plan creation process using an AI chatbot used by a company] The chatbot function provided by the AICB server 8 in this embodiment is intended to function as a "first-line contact point" for handling inquiries on the company's website, etc. This makes it possible to reduce the workload by automating inquiries that were previously handled manually.

[0129] Furthermore, the information obtained through the chatbot function will ultimately be used directly in creating the business plan.

[0130] Furthermore, the storage 84 managed by the AICB server 8 has pre-loaded fixed information such as the company's website URL, customer information, product and service information, and founding history.

[0131] As a result, the AICB server 8 automatically generates anticipated questions and answers from customers operating the user terminal 3.

[0132] Furthermore, the AICB server 8 can acquire and analyze the chatbot's history. In this context, inquiries from prospective customers (regarding price, delivery time, differentiating points, etc.) are considered "offers" indicating interest in the product or service, and the AICB server 8 manages them as extremely important data.

[0133] The AICB server 8 then acquires the conversation history with prospective customers obtained via the chatbot and manages and analyzes it in conjunction with the CRM. It also stores data such as the content of the chatbot inquiries, user information, the status of the conversation, and the circumstances leading to lost deals or requests for quotes in the database.

[0134] For example, it becomes possible to present a new service on a chatbot before its official release and conduct a proof-of-concept (POC) to test market reactions.

[0135] Specifically, by incorporating the following chatbot function into business plan creation system 1, if positive responses such as "Tell me more" are received from customers, it can be considered as evidence to promote commercialization.

[0136] Note that in order for the chatbot to automatically generate questions and answers, the user needs to input the information source as part of the initial setup.

[0137] Here, the user operating user terminal 3 registers links to web pages containing company information, product information, etc., with the AICB server 8. Currently, up to three web page links can be registered with the AICB server 8, but this number can be increased as needed.

[0138] Furthermore, even if the information provided by the user terminal 3 on the network 4 is scattered across multiple pages, the AICB server 8 can read the information by the user appropriately entering those links.

[0139] The chatbot function provided by Business Plan Creation System 1 can autonomously learn and improve upon "good conversations" that lead to conversions.

[0140] First, the AICB server 8 can be configured to automatically generate sets of anticipated questions and answers from information such as web pages published by companies, and to automatically replace conversations deemed to have poor performance (for example, the worst 20%) with better ones.

[0141] On the other hand, this system also manages information (non-structured data) received as feedback from customers via the chatbot function as an evaluation metric.

[0142] For example, based on the chatbot's responses, this system suggests to user terminal 3 that conversations that the customer has clearly rated as "Bad" be replaced with other undesirable conversations.

[0143] Similarly, if the chatbot's response indicates that the customer answered a question but the conversation ended without continuing, it can be determined that the answer was insufficient.

[0144] This allows companies' chatbot services to dynamically generate responses based on context, rather than always returning fixed answers.

[0145] Specifically, when a user asks a vague question like "price," the system extracts relevant answers from a pre-prepared set of anticipated questions and answers, combines (merges) them, and generates a single answer.

[0146] Because the answers are not fixed, responses not included in the anticipated Q&A set may be generated. The evaluation of the generated responses depends on the subsequent development of the conversation. If the conversation continues based on the responses generated in this way and leads to more specific questions (for example, "How much money is needed to create a business plan?"), the response generation process will be evaluated as "good."

[0147] The AICB server 8 can store this series of conversations as data, for example in storage 84, and analyze it to identify which question-and-answer combinations contributed to the conversion.

[0148] Furthermore, the chatbot function service can also be provided as a function of server 5 shown in Figure 29.

[0149] [Example of dialogue evaluation and scoring process] The CB management unit 11B of this system analyzes the content of the dialogue in detail, with the chatbot dialogue history stored in the database, and provides a process to quantify (score) the quality of the dialogue and the evaluation of the new service.

[0150] Here, the scoring criteria are based on past success stories (e.g., conversations that led to a sale, conversations that resulted in a request for a quote). In other words, by comparing the conversation patterns of success stories, we can evaluate the effectiveness of individual conversations (such as the keywords they contained; for example, if the phrase "That's amazing" is used, a higher number of "amazing"s will result in a higher conversation score).

[0151] On the other hand, the CB Management Department 11B can suggest modifications and improvements to question-and-answer pairs that receive low scores, so as to better meet the customer's intent. For example, in response to the question, "Is it free to use?", instead of simply offering a free plan, the quality of the conversation can be improved by providing additional information (offers) such as details of paid plans and contract types. Here, the evaluation criteria are set so that the score fluctuates depending on the purpose of the conversation and is not treated as a fixed value.

[0152] In this system, the term "score" as an evaluation metric for conversations is not simply a numerical score, but rather an indicator used to determine whether a conversation is leading to a conversion (such as obtaining an inquiry or request for a quote).

[0153] Furthermore, the definition of a conversation that leads to conversion is that specific requests from customers, such as "Please tell me," "Can you do that?", and "Please give me a quote," are positioned as important conversations (love letters) that lead to conversion. More specifically, the indicator will evaluate the qualitative flow of the conversation, rather than just a numerical score, to see if it is moving towards the final goal (e.g., requesting a brochure, requesting a quote). Just as a good salesperson can draw out all of the customer's needs and make a great proposal, the AI ​​chatbot will also consider conversations that draw out the customer's needs to the fullest extent as the "correct" answer and will give it high marks.

[0154] [Business Plan Creation Process] As the primary point of contact with customers, this system provides tools to automate a company's profit-generating activities (from acquiring prospective customers to closing deals). Therefore, the CB Information DB 11A is equipped with storage 64 for accumulating data in the CRM.

[0155] Specifically, the system employs a configuration in which customer inquiries from user terminal 3, responses to those inquiries, and concrete results such as "Please give me a brochure" or "Please give me a quote" are all stored as records in storage 64 of the CRM.

[0156] This allows the Management Analysis Department 16 to visualize and graph the conversion rate from the above-mentioned inquiries to quotations.

[0157] Specifically, the Management Analysis Department 16 acquires three types of data as input information. The three types of data are: (1) Fixed Q&A information from the initial setup, which includes basic company information, philosophy, history of independence, and other information related to the core of the business. (2) CRM data acquired from the AI ​​chatbot, which includes the history of conversations with customers and information on proof of concept obtained through the AI ​​chatbot. (3) Financial statement information, which includes information uploaded from the user terminal 3 in PDF or other format, such as a complete set of tax returns (BS / PL, details of selling, general and administrative expenses, etc.).

[0158] [Flow from analysis to business plan creation] When the user presses the "Management Analysis Button" displayed on the display unit 32 of the user terminal 3, the system 1 reads the information from (1) to (3) above, and the management analysis unit 16 automatically starts the analysis.

[0159] The Management Analysis Department 16 conducts its analysis according to fixed headings (keywords) such as "philosophy," "added value," and "customers."

[0160] In this case, if the management analysis department 16 finds that there is insufficient information during the analysis process, the AICB server 8 will automatically generate additional questions based on CRM data and publicly available information, and perform processing to improve the accuracy of the analysis.

[0161] In this system, AICB Server 8 analyzes lengthy documents such as the application guidelines (approximately 60 pages), determines whether the content is relevant to answering the business analysis questions, and extracts the information. AICB Server 8 then compares the extracted information with the review criteria required by the subsidy application documents, and examines the content while organizing it into a business plan.

[0162] In this system, commands to the AICB server 8 are standardized, and documents are read in blocks such as pages. The system automatically creates headings that align with the themes requested by the management analysis department 16 and generates the content.

[0163] This eliminates concerns that users operating user terminal 3, who are unfamiliar with business analysis, may be satisfied with the initial analysis results generated by the AICB server 8 and not continue with detailed discussions.

[0164] Furthermore, one of the functions provided by this system is to encourage user feedback and improve the accuracy of analysis by having the business plan creation system 1 present "unanswered" items to the user terminal 3.

[0165] Furthermore, once a customer's application documents pass the approval review process through this system—that is, after the subsidy has been approved by the recipient—it is possible to incorporate a function into the system that defines requirements based on the approved business plan and investment amount, and creates a request for quotation and requirements specification document.

[0166] This allows us to extract investment amounts and requirements for new businesses from the business plan and provide customers with an automated service to generate draft cost estimates.

[0167] [Example of Business Plan Creation Process] Figure 30 is a flowchart illustrating the operation of an AI chatbot according to one embodiment of the present invention. This process is based on the steps shown in Figure 10 and is characterized by the inclusion of chatbot function processing.

[0168] In the AI ​​chatbot processing in step S312, as shown in Figure 30, the server 5 and the AICB server 8 communicate via API (steps S211 and S311), and an answer is generated from the AICB server 8 using the learning model 2C in response to the user's question input, and this is displayed on the user terminal 3. Next, the user interface screen shown in Figure 29 is displayed on the display unit 32 of the user terminal 3, and the user presses the business analysis button (S112). When the server 5 detects that this business analysis button has been pressed (S212), the chatbot unit 11C executes AI chatbot processing (S312) and performs business analysis processing based on unstructured data acquired and stored through question and answer with the user (S213).

[0169] Meanwhile, if the AICB server 8 determines that it has all the necessary non-standard data (data) for business analysis obtained through question-and-answer sessions by the chatbot unit 11C (S313), it transmits the customer information to the business analysis unit 16 (S314).

[0170] Furthermore, evaluation input is performed by the user, and scoring is executed (step S111). The above process is repeated unless there is an instruction to the TOP screen (NO route of step S113).

[0171] Then, when the user input is received and the button 93A returns to the TOP screen 93 is pressed, the processing of the user terminal 3 returns to step S102 described above (YES route of step S112).

[0172] Then, the connection between server 5 and AICB server 8 is terminated, and the termination process is executed (steps S214 and S315). At this time, server 5 returns to the process of step S201, and AICB server 8 enters a waiting state until the next connection.

[0173] Regarding the AI ​​chatbot, the history of questions from each user is registered in the CB information DB 11A, linked to the access date and time. As shown in Figure 11, the operator of Server 5 can display a list of the history shown in the CB information DB 11A on the display 54.

[0174] Furthermore, the operator of server 5 associates the answers with each question and registers them in the CB information DB 11A along with the data of the last updated date and time. The operator of server 5 can display the question and answer sets on the display 54, as shown in screen 94 in Figure 12, and can also modify them. The back button 93A remains displayed.

[0175] Furthermore, the commands to the AICB server 8 are standardized, and the system reads documents in blocks such as pages, automatically creates headings aligned with the business analysis theme, and generates the content.

[0176] This is because users unfamiliar with business analysis may be satisfied with the initial analysis results generated by the AI ​​and may not continue with a detailed discussion.

[0177] The MA server 7 may also take an approach to improve analysis accuracy by prompting user feedback by presenting "unanswered" items to the user terminal 3.

[0178] According to this embodiment, by linking multiple information analysis server devices, even those who are not good at understanding the application guidelines for subsidies or the systems for fundraising through loans or equity can automatically generate a business plan that includes customer information and management analysis information required by the subsidy recipient, in accordance with the secretariat's review criteria, while already having an analysis of the business's management situation.

[0179] [Example of operation based on the user interface screen] Figures 31 to 40 show examples of user interface screens displayed on the display unit 32 of the user terminal 3 shown in Figure 1.

[0180] Figure 31 shows the dashboard screen of the subsidy application operation screen that the server 5 presents to the user terminal 3. Initial settings include common items, business plan creation support, AI chatbot, customer management / sales promotion system, subsidy information delivery, and performance reporting items, with performance reporting being configurable.

[0181] Furthermore, the system is configured so that users can select each of the above items to display them on the corresponding hierarchical operation screens.

[0182] Furthermore, this screen operation assumes that the user account has been registered on the registration screen provided by the server 5 by operating the user terminal 3.

[0183] [Basic Information Registration Process] When the "Common Items" option is selected in the screen shown in Figure 31, the basic information registration screen (Figures 32 to 34) is displayed on the display unit 32 of the user terminal 3 from the server 5. Here, the user can register the company's address and website link in the "Your Company" item to the server 5.

[0184] Furthermore, in the "Customer" section, users can register customer information on Server 5 by entering the customer's industry / business type (major category), such as "information and communications industry," and their industry / business type (medium category), such as "information services industry."

[0185] Furthermore, users can register information about their trading partners on Server 5 by entering their industry and business type (subcategory), such as software industry, and their industry and business type (detailed category), such as packaged software industry.

[0186] Next, in the "Company Concept" section, users can register with server 5 by entering links to websites related to their management philosophy and products / services.

[0187] From this point onward, 16 items are set as pre-registered questions from the chatbot unit 11C to the server 5.

[0188] As shown in Figure 33, for example, the first preliminary question is, "How did your company's representative start the company? Please tell us about the background of its founding."... The 16th question is, "Please tell us about any difficulties your company is facing in terms of sales promotion, including specific examples." Non-standard data-based questions such as these are registered in the CB information DB 11A in an updatable manner.

[0189] Therefore, the user of user terminal 3 can, simply by answering the first to sixteenth questions generated by the chatbot unit 11C presented by server 5, have information about the company's founding and challenges gathered and registered as customer information linked to the customer ID in the CB information DB 11A.

[0190] In this example, when the server 5 receives a response from the user operating user terminal 3 stating, "We haven't found a strategy to ensure our marketing measures reach our target customers," the analysis target is narrowed down to the management analysis department 16.

[0191] Next, as itemized financial information, users can upload financial statements (files containing financial indicator information). Users can upload the current period, previous period, and the period before that to server 5.

[0192] Next, the server 5 presents a screen for the user operating the user terminal 3 to select the subsidy application business plan menu, as shown in Figure 35.

[0193] In this example, an AI generation button and a free LLM (Licensed Liberal Marketing) business analysis button are provided.

[0194] Since Server 5 has already selected the appropriate application destination for the company currently accessing User Terminal 3 from the publicly available subsidy application destination information based on the Q&A with User Terminal 3, clicking "Select a subsidy" will display the subsidy application chart screen shown in Figure 36 from Server 5 to User Terminal 3.

[0195] Currently, the chart shows that, as part of the procedure, phases 1 to 3 of the seven progress screens have been completed, and the user is now in the phase where they need to input the company profile. At this point, the user on user terminal 3 enters links to the company's management philosophy website and its products and services website as part of the company profile, and then clicks the "Next" button.

[0196] As a result, the pre-question first screen is displayed on the display unit 32 of the user terminal 3 by the server 5.

[0197] Here, the chatbot unit 11C responds to question 1 (where the question corresponds to the answer item requested by the grant application recipient) with the user's answer: "By obtaining a degree in natural language processing and working in customer management systems as a company employee, I acquired the skills of dynamic information management and built my own customer management system and marketing automation."

[0198] Next, the chatbot unit 11C responded to question 2 with the statement, "I felt that creating a business plan could serve as a compass for my business, and therefore I am very interested in deepening my knowledge of how to create highly accurate business plans." Questions 3 and 12 are omitted below.

[0199] The chatbot unit 11C obtains the response "Due to my low credit score, I will not be able to do business with companies that evaluate based on company size and capital strength" from the user terminal 3 in response to question 13.

[0200] Similarly, the chatbot unit 11C obtains the response "It is difficult to raise the unit price of transactions because our main business is with small and medium-sized enterprises" from the user terminal 3 in response to question 14. When the user clicks the next button, the server 5 displays the screen showing Phase 6 (Financial Information) as shown in Figure 39 to the user terminal 3.

[0201] Here, user terminal 3 selects the previous year's tax return and uploads the financial information to server 5 in a specified file format. Furthermore, when the user clicks the "Next" button, server 5 displays the Phase 7 screen shown in Figure 40 to user terminal 3.

[0202] Figure 41 shows an example of the management philosophy information generation screen displayed on the display unit 32 of the user terminal 3 shown in Figure 1. Figure 42 shows an example of the instruction screen for creating a business plan for subsidy application displayed on the display unit 32 of the user terminal 3 shown in Figure 1. The following describes an example of the business plan creation process for subsidy application in this system.

[0203] Here, the user operating user terminal 3 can choose to either press the "Start Business Analysis" button or perform business analysis using the free LLM (Lawyer-Led Management) tool.

[0204] For example, if the "Start Business Analysis" button is selected, the business analysis unit 16 generates a business philosophy to be responded to the selected subsidy application recipient, as shown in Figure 41, based on the customer information stored in the CB information DB 11A and the information that the user has answered sequentially from Phase 1 described above.

[0205] When the user terminal 3 selects the business plan creation support shown in Figure 31, the screen shown in Figure 42 is displayed by the server 5. On this screen, the user can choose whether to create a business plan for subsidy application based on AI generation or by performing business analysis using the free LLM (Lawyer-Led Management) tool.

[0206] For example, if you choose the option based on AI generation, you will be providing a paid service, and since the system is based on the premise of generating revenue through billing, a detailed business plan for subsidy applications will be generated.

[0207] On the other hand, when using the free LLM for business analysis, the system generates a business plan for subsidy applications, focusing on the general response items required by the granting organization, based on the information the user provides in response to the questions.

[0208] [Effects of the Second Embodiment] According to this embodiment, in the subsidy application business plan service, the chatbot can be made to behave like a concierge interacting with the user. The complex subsidy application document creation process can be easily simplified by extracting non-standard data obtained from the chatbot's question-and-answer session, such as the company's founding history, and automatically generating compelling documents that fit the required format of the subsidy application and are more likely to pass the review process. In other words, a smooth conversation can be expected that cannot be obtained with so-called template-type formulaic responses, and even staff members unfamiliar with subsidy applications can respond naturally, allowing them to obtain and scrutinize non-standard data to be included in the subsidy application documents.

[0209] Furthermore, the chatbot will behave like a concierge to the user, providing flexible and accurate questions that are relevant to the context. This will enable smooth and natural responses that cannot be obtained with so-called template-based, pre-defined questions, and will allow the system to acquire non-standard data necessary for subsidy applications from users.

[0210] Although each embodiment has been described in detail above, the invention is not limited to any particular embodiment, and various modifications and changes are possible within the scope described in the claims. Furthermore, it is possible to combine all or more of the components of the embodiments described above.

[0211] Furthermore, the configurations of the various functions, processes, and databases described above are merely examples. The configurations of the various functions, processes, and databases can be changed to the optimal configuration from the standpoint of the performance, processing efficiency, and communication efficiency of the hardware and software provided by these devices.

[0212] Furthermore, the configuration of the database (schema, etc.) used to store the various types of data mentioned above can be flexibly modified from the perspective of efficient resource utilization, improved processing efficiency, improved access efficiency, and improved search efficiency.

[0213] Specifically, the subject of the system or method in this disclosure is a computer.

[0214] The functions of the subject of the apparatus or method described herein are realized by the computer executing a computer program. The computer comprises, as its main hardware configuration, a processor that operates according to the computer program. The processor may be of any type as long as it can realize its functions by executing the computer program.

[0215] A processor consists of one or more electronic circuits, including semiconductor integrated circuits (ICs, LSIs, etc.). Computer programs are recorded on non-temporary recording media such as ROMs, optical discs, and hard disk drives that are readable by a computer. Computer programs may be pre-stored on the recording media or supplied to the recording media via a wide-area communication network, including the Internet.

[0216] 1. Business plan creation system 2. Generation AI core unit 2C. Learning model 3. User terminal 5. Server 6. CRM server 7. MA server 8. AICB server 11A. CB information DB 11B. CB management unit 12. Customer management DB 13. Customer management unit 14A. Application document DB 14B. Application management unit 15. Business plan creation / review unit 16. Management analysis unit 17. Application management DB 18. Output conversion unit 19. Application procedure unit 20. Operation unit

Claims

1. A business plan creation system comprising: a user terminal that instructs the automatic generation of a subsidy application form to be submitted to a designated subsidy application recipient; and a server device that provides a service for creating a business plan for fundraising in conjunction with a learning model that holds parameters learned from response information to the items required by the subsidy application form, wherein the server device includes a database that stores the application guidelines for business plans required by the designated subsidy application recipient and customer management information of users of the service; a business analysis unit that refers to a set of anticipated questions and answers combining questions and answers regarding fundraising, uses the learning model to obtain answers to the questions from the user terminal, and performs business analysis using the learning model based on the questions and answers; and a creation unit that instructs the learning model to create a business plan using the customer management information, the application guidelines, the results of the business analysis, and an evaluation score for the answers, using a chatbot function with the user terminal, and obtains the business plan created by the learning model based on the parameters.

2. A business plan creation system according to claim 1, wherein the business analysis unit generates question items necessary for the business analysis, and performs the business analysis based on non-quantitative data generated by obtaining the answers to the question items from the user terminal via the chatbot unit.

3. A business plan creation system according to claim 1, characterized in that the database includes customer management information relating to CRM (Customer Relationship Management).

4. A business plan creation system according to claim 1, wherein the database includes the customer management information relating to MA (Marketing Automation).

5. A business plan creation system according to claim 1, wherein the creation unit creates a table corresponding to each item of the business plan, the customer management information stored in the database and the analysis targets of the management analysis unit, and provides this table to the user terminal, and accepts from the user terminal the selection of the items to be reviewed.

6. A business plan creation system according to claim 5, wherein the creation unit modifies the information corresponding to the selected item from among the customer management information stored in the database and the analysis targets of the management analysis unit in accordance with the operation of the user terminal.

7. A business plan creation system according to claim 1, further comprising an output unit that files the business plan created by the creation unit and outputs it to the user terminal.

8. A business plan creation system according to claim 1, wherein the database is characterized in that it adds and stores unstored application guidelines in response to the operation of the user terminal.

9. A business plan creation system according to claim 1, further comprising an application procedure unit that is connected to a server to which applications for funding are submitted, and which submits an application to the server using the business plan created by the creation unit.

10. A business plan creation system according to claim 1, characterized in that the funding is a subsidy, grant, loan, or investment.

11. A business plan creation system according to claim 1, characterized in that the items of the business plan include one or more of the following: company overview, background, management challenges, market analysis, competitor analysis, analytical methods, and identification of challenges.

12. A business plan creation system according to claim 1, characterized in that it comprises a management unit that manages the evaluation of the responses using scores.

13. A business plan creation system according to claim 1, characterized in that the learning model holds the following parameters: approved subsidy applications from the application destination, rejected applications from the application destination, application guidelines, application criteria specific to the application destination, information on the company applying for the subsidy, and review comments for reviewing the subsidy application.

14. A business analysis method for a business plan creation system, comprising a user terminal that instructs the automatic generation of a subsidy application form to be submitted to a designated subsidy recipient, and a server device that provides a function to create a business plan for fundraising and a chatbot function by linking a user terminal that instructs the automatic generation of a subsidy application form to be submitted to a designated subsidy recipient, the server device having a database that stores the application guidelines for business plans requested by the designated subsidy recipient and customer management information of the service users, and a business analysis step that refers to a set of hypothetical questions and answers combining questions and answers regarding fundraising, obtains answers to the questions from the user terminal using the learning model, and performs business analysis using the learning model based on the questions and answers, and a creation step that instructs the learning model to create a business plan using the customer management information, the application guidelines, the results of the business analysis, and an evaluation score for the answers, using prompts input from the user terminal, and obtains the business plan created by the learning model based on the parameters, The aforementioned business analysis step is a business plan creation system that analyzes the conversation history exchanged between the customer and the client using the chatbot function.

15. A computer program that creates a business plan for fundraising in conjunction with a learning model that holds parameters learned from response information to non-standard data items required by a subsidy application, the program comprising: a business analysis step in which the computer refers to a set of hypothetical questions and answers combining questions and answers related to fundraising, uses the learning model to obtain answers to the questions from a user terminal, and performs a business analysis using the learning model based on the questions and answers; and a creation step in which the learning model is instructed to create a business plan using customer management information, application guidelines, the results of the business analysis, and evaluation scores for the answers, using prompts input from the user terminal, and obtains the business plan created by the learning model based on the parameters.

Images