system

Abstract

Provide a system. An information storage means for storing market-related information or business-related information; a prompt generating means for generating a prompt sentence instructing the user to generate strategic information regarding the establishment or growth of a business entity based on the market-related information or the business-related information stored in the information holding means; a generation means for inputting the prompt sentence into a generation AI model that performs natural language processing, and acquiring from the generation AI model at least one of new business proposal information, progress management information, fundraising support information, sales promotion information, development process improvement information, and risk avoidance information for the business entity; a presentation means for presenting the information acquired by the generation means to a user; A system including:

Description

[Technical Field]

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

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

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In the past, it was difficult to smoothly obtain information on new business proposals, progress management, fundraising support, sales promotion, development process improvement, or risk avoidance during the process of establishing or growing a business. As a result, businesses faced a lack of information to make appropriate strategic decisions, which could lead to delays in growth and increased risks. This required a great deal of effort on the part of those involved to search for and analyze information, making it difficult to execute business plans quickly and effectively. [Means for solving the problem]

[0005] In this invention, the server includes an information storage means, a prompt generation means, and a generation means, which allows a business entity to quickly obtain necessary strategic information through prompt sentences input to a generative AI model based on market-related information or business-related information.

[0006] The "information storage means" is a functional unit that stores market-related information or business-related information relating to a business entity in a storage device or recording medium and stores the information in a state that can be referenced in subsequent processing. "Market-related information" refers to information that indicates the trends of the target market, customer preferences, the competitive situation, etc., and is data that a business entity should refer to when making strategic decisions. "Business-related information" refers to various data necessary for making decisions regarding business operations, such as the business entity's internal plans, financial status, development progress, organizational structure, etc. An "entity" is a newly established organization or an entity that operates a specific business and aims to grow and develop. "Strategic information" refers to any of new business proposal information, progress management information, fundraising support information, sales promotion information, development process improvement information, or risk avoidance information generated based on market-related information or business-related information, and serves as a guideline used by an entity to make strategic decisions. The "prompt generation means" is a functional unit that constructs a prompt sentence to be input to the generative AI model based on the market-related information or the business-related information. A "prompt sentence" is a query or instruction sentence in natural language format that is input to a generative AI model, and is input data that causes the model to generate desired strategic information. A "generative AI model" is a trained model that uses natural language processing technology to automatically generate strategic information useful to a business entity in response to an input prompt. The "generation means" is a functional unit that inputs the prompt sentence into the generation AI model, acquires the strategy information obtained as a response, and structures or processes it. The "presentation means" is a functional unit that displays or outputs the strategy information obtained by the generation means in a format that can be understood by the user. "Sales promotion information" is an example of the strategic information, and is information that includes suggestions and instructions regarding sales strategies, advertising methods, customer acquisition strategies, etc. that are suited to market trends. The term "latent factors" refers to elements or conditions that are risk sources extracted from the market-related information or business-related information, and are taken into consideration when deriving risk aversion information. "Risk prevention information" is strategic information that identifies potential problems that an entity may face in advance and includes measures to mitigate or avoid them. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 2 is a sequence diagram showing the flow of processing of the system in this embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0009] <System configuration>

[0010] An example of an embodiment of the present invention will be described below. Note that the present invention is not limited to the embodiment described below, and various changes and modifications are possible without departing from the technical concept of the present invention.

[0011] This embodiment relates to a system that enables a business entity to acquire strategic information based on market-related information or business-related information and utilize the strategic information. Such a system includes at least a server and a terminal.

[0012] (Overall system configuration) The system of this embodiment mainly comprises a server and a terminal capable of displaying information provided by the server. The server has a functional unit including an information storage means, a prompt generation means, and a generation means. The terminal has a function of visually or audibly presenting information obtained from the server, and has an interface for issuing requests to the server by a user performing an input operation.

[0013] (Server configuration) A server has a central processing unit (CPU), main memory, auxiliary storage devices (hard disk drives, solid-state drives, etc.), a network interface, and an operating system that controls these devices. The server is equipped with a program execution environment (for example, a scripting language execution environment or a web application server) and processes information in response to user requests.

[0014] The information storage means includes a database or memory area for storing market-related information or business-related information. Market-related information includes market trends, customer preferences, competitive situations, fundraising trends, etc., while business-related information includes an entity's internal plans, financial status, development progress, organizational structure, etc. The database can be any data storage format, such as a relational database management system or a NoSQL database system.

[0015] The prompt generation means dynamically generates prompt statements to be input into the generative AI model by extracting, formatting, and supplementing market- or business-related information obtained from the information storage means as needed. For example, the server analyzes a request from a business entity (such as a new business proposal, progress management improvement, fundraising support, sales promotion measures, development process improvement proposals, or a request for the presentation of potential risk countermeasures) and selects the information necessary to satisfy the request from the information storage means. The server then converts and integrates the selected information into natural language text and generates an instruction statement such as, "Present a new business proposal based on the following information." This is the prompt statement, and the server can input this prompt statement into the generative AI model to obtain strategic information from the model.

[0016] The generation means has a function of inputting the prompt sentence generated by the prompt generation means into a generative AI model (e.g., a large-scale language model) and obtaining natural language text generated as a response from the model. The generative AI model can return useful strategic information corresponding to the prompt sentence based on the learning data. The generation means processes the obtained strategic information into a format that is easy for the user to read and understand (e.g., JSON format or HTML format) and transmits it to the terminal described below.

[0017] (Device configuration) A terminal is a device equipped with a display device and an operation interface (keyboard, mouse, touch panel, or voice input means) that can communicate with a server. For example, a terminal may be a personal computer, a tablet-type information processing device, or a mobile information terminal. The terminal has client software such as a web browser or dedicated application, and displays information received from the server on the screen. The user can use the terminal to check the displayed content and send a new request (for example, "Show me more detailed sales promotion ideas") to the server.

[0018] (Processing flow) When a user inputs a request from a terminal (e.g., "Present an idea that will lead to sales promotion of an existing product"), the terminal sends this request to the server. The server analyzes the request, obtains market-related or business-related information from the information storage means, and processes it using the prompt generation means. This results in a prompt sentence suitable for the generative AI model. The server sends the prompt sentence to the generative AI model via the generation means, and the strategic information obtained as the model response is reformatted again on the server side before being sent to the terminal. The terminal displays the received strategic information in a format that is easy for the user to understand. The user can issue the next request while referring to the displayed results, and the system can repeatedly provide information in response to this.

[0019] (Variation) In this embodiment, an example has been shown in which the server integrally comprises the information storage means, prompt generation means, and generation means, but these functions may also be physically distributed. For example, a database server equipped with the information storage means and an application server equipped with the prompt generation means and generation means may be configured to link over a network. Furthermore, the terminal may use a specific GUI layout or dialogue format, or may combine information presentation means using voice output or a wearable device.

[0020] As described above, according to the embodiment of the present invention, strategic information required by a business entity (new business proposal information, progress management information, fundraising support information, sales promotion information, development process improvement information, or risk avoidance information) can be smoothly and quickly obtained using a generative AI model and presented to a user via a terminal. This makes it easier for the business entity to make strategic decisions based on multifaceted information, effectively promoting the establishment and growth of the business.

[0021] <Example of system configuration>

[0022] An example of an embodiment of the present invention will be described below. Note that the present invention is not limited to the embodiment described below, and various changes and modifications are possible without departing from the technical concept of the present invention.

[0023] This embodiment relates to a system that enables a business entity to acquire strategic information based on market-related information or business-related information (hereinafter collectively referred to as "related information") and utilize the strategic information. Such a system has a configuration including at least a server and a terminal, and each of the components will be described in detail below.

[0024] (server) The server has functional units including an information storage means capable of referencing related information, a prompt generation means that uses this related information to create a prompt sentence to be input to the generative AI model, and a generation means that inputs the prompt sentence into the generative AI model and obtains strategic information from the obtained model response.

[0025] Hardware configuration: A server comprises a central processing unit (CPU), main memory (RAM), secondary memory (HDD, SSD, etc.), network interface (wired LAN, wireless LAN, optical communication line connection module, etc.), and basic input / output devices (console screen, keyboard, mouse, etc.). These hardware resources operate under an operating system (Linux®, UNIX®, Windows® Server, etc.) and provide an environment for running server programs. It is also desirable for the server to be equipped with multiple parallel processing units (GPU, TPU, etc.), which improves the computing power required for accessing generative AI models and large-scale analysis of related information.

[0026] Software configuration: The server is equipped with a database management system (DBMS) and search engine for managing related information. The DBMS can be a relational database (MySQL, PostgreSQL, etc.) or a non-relational database (NoSQL database, document-oriented data store, etc.). Related information stored in these databases includes market trend data, customer behavior analysis data, development progress data, financial planning data, potential risk data, advertising effectiveness analysis data, and competitive product information. The server also implements an API endpoint using a web application framework (Flask, Django, FastAPI, Spring Boot, etc.) to enable it to accept requests from terminals.

[0027] Information retention means: The information storage means includes the above-mentioned DBMS, file storage, and cache systems (Redis, Memcached, etc.), and stores related information permanently or temporarily. This allows the server to quickly reference the necessary information (e.g., statistics on the target market and customer attribute information) when requested by a terminal.

[0028] Prompt generation method: The prompt generation means is a functional unit implemented by a program that formats relevant information as text and assembles a prompt statement to be input to the generative AI model. Specifically, the server analyzes a request received from a terminal (e.g., "Request for presentation of new business proposal information") and obtains from the information storage means the relevant information needed to fulfill the request (market data, business plan information, customer analysis results, etc.). The server then assembles this information in a natural language format to generate an instruction statement such as "Based on the following market data, present three new business proposals." This prompt statement is input to the generative AI model, providing instructions for the model to generate appropriate strategic information.

[0029] Generation means: The generation means has the function of sending the prompt sentence obtained by the prompt generation means to the generated AI model (either an external model hosting service or a model running on an inference environment installed on the server itself) and obtaining the model response. The natural language text obtained as the model response includes strategic information such as new business proposal information, project progress management information, fundraising support information, sales promotion information, development process improvement information, or risk avoidance information. The generation means converts this model response text into a format (JSON, XML, HTML, Markdown, etc.) that is easy for the terminal to process and display, and returns it to the terminal over the network.

[0030] (Terminal) A terminal is an information processing device used by a user to exchange information with a server, and examples of the terminal include desktop personal computers, laptops, tablet devices, smartphones, and wearable devices.

[0031] Hardware configuration: The terminal is equipped with a display device (such as a display or head-mounted display) and input / output devices for operation (such as a keyboard, touch panel, microphone for voice input, click button, gesture sensor, etc.) The terminal can communicate with the server via a wireless LAN, cellular network, or wired LAN interface.

[0032] Software configuration: The terminal has a web browser or a dedicated client application and sends HTTP requests to the API provided by the server. In addition, by using a user interface (UI) front-end framework (React (registered trademark), Vue.js (registered trademark), Angular (registered trademark), etc.) that uses a scripting language such as JavaScript (registered trademark), HTML, CSS, etc. on the terminal side, the response data (JSON, HTML, etc.) from the server is analyzed and displayed as an appropriate UI component on the screen.

[0033] Display and operation interface: The terminal displays the strategic information received from the server (for example, three new business ideas) in a format that is easy for the user to understand. The user refers to the displayed strategic information, enters additional information requests (for example, "Analyze this idea in more detail" or "Show potential investors related to this idea") in the input fields on the terminal, and sends the new request to the server by pressing the send button.

[0034] Processing flow on the device: A series of processes is repeated: the user inputs a request from the terminal → the terminal sends the request to the server → the server analyzes the request, obtains related information, generates prompts, queries the model, and formats the results → the terminal displays the received strategic information on the screen. This allows the user to obtain, verify, and evaluate business strategic information in a step-by-step and dynamic manner through multiple interactions.

[0035] (effect) With the above configuration, users (business entity personnel) can simply operate their devices to sequentially obtain advanced strategic information (generated by the generative AI model) based on related information held by the server. This eliminates the need to manually perform the conventional process of extensive market research, competitive analysis, fundraising planning, risk identification, and countermeasure considerations in separate steps. This system provides strategic information systematically and immediately, enabling efficient and rapid decision-making by business entities and effectively supporting their growth and development.

[0036] <System Operation>

[0037] The following is an embodiment showing the program processing of the system according to the present invention.

[0038] In this embodiment, the server runs on a Linux (registered trademark)-based operating environment and accesses the generative AI model using high-performance computing resources (a cloud computing environment or a virtual machine equipped with a GPU). The server runs a web application server written in Python (registered trademark) (e.g., Flask (registered trademark) or FastAPI (registered trademark)). The server stores related information using a database management system such as MySQL (registered trademark) or PostgreSQL (registered trademark), and retrieves, filters, and aggregates this information using a database query language (SQL). The server converts the retrieved related information into natural language text and can construct prompt statements using a template engine such as Jinja2 (registered trademark).

[0039] The terminal is an information processing device used by the user, and uses a web browser (such as Google Chrome (registered trademark) or Mozilla Firefox (registered trademark)) to send HTTP requests to API endpoints published by the server. The terminal dynamically manipulates page elements using JavaScript, obtains the information requested by the user from the server, and displays it on the screen. The user can enter forms and operate buttons on the terminal's browser, and obtain strategic information generated by the server based on the input content.

[0040] The server performs text analysis on requests sent by users (for example, input along the lines of "Propose a strategy for developing a new market") to search for relevant information and generate prompts to send to the generative AI model. The server retrieves the business's market data and customer trend data from a database, filtering and preprocessing the relevant data using Python scripts. For example, it reads the relevant data in JSON format, deletes some unnecessary fields, converts the integrated data into natural language, and then uses a template engine to assemble prompts such as, "Based on the following market data, propose three innovative business ideas for new startups. Market data: [Customer behavior analysis results... List of competing services...]."

[0041] The server sends this prompt as an HTTP request to a generative AI model (e.g., OpenAI's (registered trademark) GPT model or other large-scale language model). The generative AI model returns the desired strategy information in response to the prompt, and the server formats the response into a general-purpose data format such as JSON. When the device receives the formatted information, it displays it on the user interface through DOM manipulation using JavaScript. The user can refer to the displayed information and input additional requests. Through these successive exchanges, the user can efficiently obtain the desired strategy information.

[0042] <Example of system operation (Figure 1)>

[0043] Step 1: The user inputs the request content "Show me a new business proposal" as text into an input field on the terminal, and sends this request content to the terminal by pressing the send button. (Input) User input text: "Show us your new business proposal." (Output) Request data passed to the terminal (user request)

[0044] Step 2: The terminal converts the request data received from the user into a request message in JSON format and sends an HTTP request to the server. (Input) User request data ("Show us your new business proposal") (Output) Request message to be sent to the server (JSON format)

[0045] Step 3: The server analyzes the request message received from the terminal and retrieves market-related information and business-related information from the information storage means (database) using an SQL query. The server converts the retrieved data into natural language text and generates a prompt sentence based on that text using a template engine. For example, "Propose three innovative business ideas for a new startup based on the following market data: [Customer behavior analysis results... List of competing services...]" The prompt sentence is constructed as follows. (Input) Request message (JSON format), market-related information and business-related information in the database (Data processing and calculation) Request analysis, database search (SQL), text formatting (removal of unnecessary elements, natural language processing), application of prompt statement templates (Output) Prompt sentence to input to the generative AI model

[0046] Step 4: The server uses the generation means to send the generated prompt sentence to the AI ​​model via an HTTP request and receives a model response. The model analyzes the prompt sentence and generates three new ideas in natural language as a response. (Input) Prompt statement (Data Calculation) Natural Language Generation Processing Using Generative AI Models (Output) Response text from the generative AI model (e.g., three new business proposals)

[0047] Step 5: The server formats the response text obtained from the generative AI model into a data structure that is easy for the device to interpret, such as JSON or HTML. The server converts the bulleted information in the text into a list structure and performs data processing such as removing unnecessary line breaks and redundant expressions. (Input) Generated AI model response text (Data processing) Text analysis, structured formatting (JSON), grammatical adjustment (Output) Formatted response message (JSON, etc.) to be returned to the terminal

[0048] Step 6: The server returns the formatted response message to the terminal as an HTTP response. (Input) Formatted response message (JSON, etc.) (Output) Response data (JSON, etc.) sent to the terminal

[0049] Step 7: The terminal uses JavaScript to analyze the response data received from the server, generate and update HTML elements, and display them on the user screen. For example, to display three new business proposals in bullet points, the terminal generates list items using DOM manipulation and presents them in a form that is visually easy for the user to understand. (Input) Response data from the server (JSON, etc.) (Data processing) DOM manipulation, HTML element generation, screen drawing processing (Output) New business proposal information displayed to the user

Claims

[Claim 1] an information holding means for holding market-related information or business-related information; performing text analysis of a request to generate new business proposal information for a business entity input by a user, searching the information storage means for the market-related information or the business-related information corresponding to the request, and converting the searched market-related information or the business-related information into natural language; a prompt generating means for generating a prompt sentence including the converted market-related information or the business-related information, the prompt sentence instructing the user to generate new business proposal information for a business entity based on the market-related information or the business-related information; a generation means for inputting the prompt sentence into a generation AI model that performs natural language processing, and acquiring from the generation AI model at least new business proposal information for a business entity; a presentation means for presenting the information acquired by the generation means to the user; A system including:

Images