Business information processing system, business information processing method, and business information processing program

Abstract

To provide a business information processing system that enables users to efficiently carry out business in a situation where a plurality of business systems exist within an organization. [Solution] The business information processing system includes a language model 11 that has been machine-learned to output a response when an output instruction is input in natural language, multiple business support systems 13, and an API 14 that is an interface to the business support systems. When an output instruction or an output request is input to the language model 11, the language model 11 selects the API 14 required for the output instruction or output in accordance with the output instruction, and sends a request to the API 14.

Description

[Technical Field]

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

[0002] 2. Description of the Related Art In companies, many types of business support systems are used in their operations, such as attendance management systems, approval systems, expense management systems, contract management systems, ordering systems, and email transmission / reception systems. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] https: / / the-board.jp / solutions / solutions_for_operation_management_system_and_erp Summary of the Invention [Problem to be solved by the invention]

[0004] When an organization has many business systems, users must consider which business systems they need to use and how to use them to carry out a certain task, which can be time-consuming and labor-intensive.In particular, when multiple business systems must be used to carry out a job, there is the problem that a lot of time is spent considering how to use multiple business systems and operating multiple business systems.

[0005] An object of the present invention is to provide a business information processing system, a business information processing method, and a business information processing program that allow users to efficiently carry out their business in a situation where multiple business systems exist within an organization. [Means for solving the problem]

[0006] In order to solve the above problem, a business information processing system according to an embodiment of the present invention comprises a language model that has been machine-learned to output an answer when an output instruction is input, a plurality of business support systems, and an API that is an interface to the business support systems. When an output instruction is input to the language model, the language model selects the API required for output in accordance with the output instruction and sends a request to the API.

[0007] The business support system corresponding to the API to which the request was sent may send an output corresponding to the request to the language model, and the language model may determine the next action based on the output sent by the business support system.

[0008] The business support system may include an approval system and an ordering system, and when an approval request is approved in the approval system, the language model may extract purchasing information contained in the approval request and input the extracted purchasing information into an API, which is an interface of the ordering system, and the ordering system may execute purchasing processing corresponding to the purchasing information input into the API.

[0009] The business support system may include an accounting system, and when the language model is instructed to compare input information regarding a voucher with information related to the voucher, the language model may send a request to the API of the accounting system including a request to output the input information regarding the voucher, compare the input information regarding the voucher output by the accounting system in accordance with the request with the information related to the voucher, and output the results of the comparison.

[0010] When the language model is instructed to compare input information regarding a voucher with information recorded in an image of the voucher, the language model may send a request to the API of the accounting system including an instruction to output the input information regarding the voucher, compare the input information regarding the voucher output by the accounting system in accordance with the request with the information recorded in the image of the voucher, and output the comparison result.

[0011] The business support system may include an invoice management system, an approval system, and a payment processing system, and when an invoice image is uploaded to the invoice management system, the language model sends a first request to the API of the approval system, including an output instruction to search for an approved approval document corresponding to the transaction recorded in the invoice image, and when the approval system outputs a response indicating that an approved approval document for the transaction exists in response to the first request, the language model sends a second request to the API of the payment processing system, including an output instruction to perform payment processing corresponding to the transaction.The system may also include a business support system function list in which functions of the business support system are recorded, and when an output instruction is input to the language model, the language model refers to the business support system function list to select an API required for output in accordance with the output instruction and send a request to the API.

[0012] A business information processing method according to an embodiment of the present invention includes an output instruction input step executed by a computer, in which an output instruction is input to a language model that has been machine-learned to output a response when an output instruction is input; a request transmission step in which the language model selects an API required for output in accordance with the output instruction from a plurality of APIs and transmits a request to the selected API; and a request execution step in which the request is executed by a business support system corresponding to the API to which the request was transmitted in the request transmission step, and the business support system outputs a response.

[0013] A business information processing program according to an embodiment of the present invention causes a computer to execute an output instruction input step of inputting an output instruction to a language model that has been machine-learned to output a response when an output instruction is input; a request transmission step of selecting an API from a plurality of APIs that the language model needs to output in accordance with the output instruction and transmitting a request to the selected API; and a request execution step of causing a business support system corresponding to the API to which the request was transmitted in the request transmission step to execute the request and output a response from the business support system. [Effects of the Invention]

[0014] It is possible to provide a business information processing system, a business information processing method, and a business information processing program that enable users to efficiently carry out their business in a situation where multiple business systems exist within an organization. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a diagram illustrating a configuration of a business information processing system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram showing the configuration of a computer used in the present system. [Figure 3] FIG. 2 is a flowchart showing information processing according to an embodiment of the present invention. [Figure 4] FIG. 10 is a diagram illustrating an example of a business support system function list. [Figure 5] FIG. 1 is a diagram illustrating a configuration of a business information processing system according to a first embodiment. [Figure 6] FIG. 10 is a diagram illustrating a configuration of a business information processing system according to a second embodiment. [Figure 7] FIG. 10 is a diagram illustrating a configuration of a business information processing system according to a third embodiment. [Figure 8] FIG. 10 is a diagram illustrating a configuration of a business information processing system according to a fourth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0016] 1 is a diagram illustrating an embodiment of the present invention. A business information processing system according to the embodiment of the present invention includes a language model 11 that has been machine-learned to output a response when an output instruction is input in a natural language, a plurality of business support systems 13, and an API 14 that is an interface for the business support systems. When an output instruction or an output request is input to the language model 11, the language model 11 selects an API 14 required for the output instruction or for output in accordance with the output instruction, and transmits a request to the API 14.

[0017] A business support system is a system or application program that supports specific operations within an organization, such as accounting, human resources, production, logistics, and sales, and examples include, but are not limited to, expense claim systems, attendance management systems, approval systems, ordering systems (including purchasing systems), contract management systems, accounting systems, payroll management systems, email sending and receiving systems, communication systems, patent information management systems, tax management systems, production management systems, sales support systems, and entrance and exit management systems. Business support systems also include ERP (Enterprise Resource Planning) systems, integrated core business systems, core systems, and business management systems.

[0018] API is an abbreviation for Application Programming Interface. An API is an interface for exchanging information between software, programs, and web services (inputting information from other software into software, etc.). Typically, users input information and input output instructions into software, but software that has an API can also input information from other software through the API.

[0019] A language model is a model that uses machine learning to learn training data (teacher data) so that when an output instruction or request in natural language is input, it outputs a response according to the output instruction, etc. Language models can process not only natural language but also programming languages. A language model constructed using a large amount of data is called an LLM (Large Language Model). In this specification, a language model with a training data capacity of 1 GB or more is called an LLM (Large Language Model).

[0020] The information processing of this system is realized by a computer. FIG. 2 is a diagram showing the configuration of a computer 20 used in the information processing of this system. The memory unit 21 is a device having the function of storing information, and in this system, a language model 11 is stored in the memory unit 21. The memory unit 11 is composed of, for example, a hard disk, an SSD, a semiconductor memory, etc. The calculation unit 22 is a device for executing a program read from the memory unit 21. The output unit 24 is a device for outputting information stored in the memory unit 21 in a manner recognizable by the user, and is composed of, for example, a display, a speaker, etc. The input unit 23 is a device for the user to input information to the computer 20, and is composed of, for example, a keyboard, a mouse, a microphone, and a camera. The communication unit 25 is a device for inputting and outputting information to and from devices outside the computer 20.

[0021] 3 is a flowchart of information processing in this embodiment. First, a user inputs an output instruction (request) to the LLM. The LLM then understands the user's request and determines the actions that must be taken to achieve the request (ST1).

[0022] Next, the LLM determines which business system with an API should be used to execute the determined action (ST2). Specifically, the LLM refers to a business support system function list (see Figure 4), which records the functions of each business support system with an API, and identifies the business support system that can execute the action based on the list. Figure 4 shows List 4, which records the main functions of the business support system. List 4 only lists the main functions of the business support system, but an actual business support system function list records the functions of the business support system in detail and comprehensively.

[0023] Then, in order to execute the determined action, the LLM creates a request including keywords for executing the action to be sent to the API corresponding to the available business support system (ST3).Then, the LLM sends the created request to the API and makes a query to the API (ST4).

[0024] The business support system corresponding to the API that received the inquiry outputs an answer (also called a reply or response) and sends it to the API of the LLM (ST5). The LLM checks the response, and if the response contains an error or is different from the expected response, it returns to ST3 (ST6), corrects the request, and sends the corrected request to the API.

[0025] If the response is as expected, it is determined whether the requirement has been met (ST7). If the requirement has not been met, the process returns to ST2 to execute the next action, and steps ST2 to ST7 are executed for the next action. If the request is met in ST7, the process ends and the answer is output (ST8). [Example]

[0026] FIG. 5 is a diagram showing the configuration of a business information processing system according to the first embodiment. In an approval system 531, a person in charge submits an approval request, and an approval authority sequentially approves or disapproves the request. The approval system 531 has an API 541 for inputting and outputting information to and from external software. An ordering system 532 has a function for searching whether a desired product is available for purchase and a function for transmitting order information to an ordering party. The ordering system 542 also has an API 542 for inputting and outputting information to and from external software.

[0027] In the approval request system 531, if the request for approval includes the purchase of a product, once the request is approved, the system is programmed to send a predetermined output instruction to the LLM's API: "Based on this request, please search the purchasing system for the target product, and purchase it if it is available. Please purchase it only if it complies with the conditions of the request. If the target product is not available, please notify the request submitter." When the above output instruction is input to the API 52 of the LLM 51, the LLM identifies an action to output in accordance with the above output instruction, and refers to the business support system function list (see Figure 4) to search for a business support system available within the organization that can execute the action. As a result of the search, the LLM identifies that the action can be executed by the ordering system, and selects the API 542 of the ordering system 532 as the destination of the request.

[0028] Next, the LLM 51 creates a request to execute the action in a format that can be processed by the API 542 of the ordering system 532, to be sent to the API 542 of the ordering system 532. The content of the request embodies the above output instructions, but because the API 542 of the ordering system 532 cannot process natural language, it must be converted into a format that the API 542 can process.

[0029] When the created request is sent to the API 542, the contents of the request are executed by the ordering system 532. Specifically, it is confirmed whether the desired product described in the request can be purchased in accordance with the conditions described in the request and the company rules, and if it can be purchased, the order processing is executed, and if it cannot be purchased, the person who submitted the request is notified of this fact by email or other communication tool.

[0030] According to this embodiment, if the approval request is approved, the product purchase process is automatically carried out without the applicant having to perform any operations on the approval request system or the ordering system, which reduces the burden on the user's business processes and improves business efficiency. In addition, the necessary purchase process is reliably carried out.

[0031] In addition, because output instructions for LLM are expressed in natural language, users can easily add new output instructions according to business needs, even if they do not have programming knowledge. [Example]

[0032] FIG. 6 is a diagram showing the configuration of a business information processing system according to a second embodiment. An accounting system 631 stores data on expenses and expenditures. An invoice image management system 632 stores data on expenditures read by applying AI-OCR to invoice images. AI-OCR (Optical Character Recognition / Reader) not only extracts text information from images but also extracts text information corresponding to predetermined items.

[0033] Suppose a user inputs an output instruction to LLM 61, such as "Please tell me the patent-related expenses from January to September 2024." or "Please verify the patent-related expenses from January to September 2024." LLM 61 compares the content of the output instruction with the content recorded in the business support system function list (see Figure 4), identifies two business support systems, accounting system 631 and invoice image management system 632, as appropriate for executing the action corresponding to the output instruction, and selects APIs 641 and 642 corresponding to these business support systems as the request destinations. Verification refers to comparing two or more types of information and confirming whether the information is correct. Then, requests for each API are created in a format that can be processed by API 641 and API 642. The created requests are sent to API 641 and API 642, respectively, and the accounting system 631 and invoice image management system 632 execute the requests, respectively. As a result, if accounting system 631 outputs a response stating, "Patent-related expenses from January to September 2024 are 10 million yen," and invoice image management system 632 outputs a response stating, "Patent-related expenses from January to September 2024 are 10.2 million yen," the user will be able to recognize that 200,000 yen worth of patent-related expenses may not have been entered into the accounting system, or that an incorrect invoice image may have been stored in association with the transaction.

[0034] If a user were to operate an accounting system and an invoice image management system to conduct research, a task that would take hours would be completed. However, with this embodiment, by simply entering a question in natural language into the LLM, such as "Please tell me the patent-related expenses from January to September 2024," an answer would be output instantly, significantly improving work efficiency. Even if a question is not anticipated in advance and the system does not have a pre-programmed answer process, an appropriate answer can be obtained instantly. Furthermore, with this embodiment, even if a user does not know how to operate the accounting system or the invoice image management system, they can use the business support system to obtain the information they are looking for simply by entering output instructions in natural language.

[0035] This system can also efficiently perform a variety of other verification tasks. For example, if you want to check whether the amount of an employee's travel expense claim is appropriate, you can issue an output instruction to the LLM. The LLM will identify the action to take to respond in accordance with the output instruction, identify the expense claim system as a business support system that can execute the action, and send a request to the expense claim system's API to query it, thereby obtaining data on the employee's claim route and the travel expenses claimed. It then selects the travel expense search system as a business support system that can execute the action, and sends a request to the travel expense search system's API to query it, obtaining data on normal travel expenses in the claim route. The LLM will then compare the travel expenses claimed by the employee with the normal travel expenses in the claim route, determine whether the employee has claimed excessive travel expenses, and output the result of the determination, allowing you to instantly know whether the amount of the employee's travel expense claim is appropriate simply by making a query in natural language. [Example]

[0036] FIG. 7 is a diagram showing the configuration of a business information processing system according to a third embodiment. An email transmission / reception system 731 is an application program for transmitting and receiving emails, and has an API 732 for inputting and outputting information with external software. A tax LLM 741 is an LLM specially constructed for answering questions about taxation, and has an API 742. Although the tax LLM 741 can only output answers to questions about taxation, it can provide more accurate answers to questions about taxation than a general-purpose LLM.

[0037] Assume that a user inputs an output instruction to the LLM 71 requesting an answer to a tax-related question. The LLM 71 creates two actions in response to the output instruction: Action 1 "Contact an external tax expert" and Action 2 "Contact a tax LLM." The LLM 71 then refers to the business support system function list (see Figure 4) to identify the email transmission / reception system 731 as the business support system for executing Action 1, and selects the API 732 of the email transmission / reception system 731 as the request destination. The LLM 71 creates, as the content of the request, an email message to inquire about the appropriate external tax expert and the expert's email address, and sends the request to the API 732 of the email transmission / reception system 731. The email transmission / reception system 731 executes the request and sends an email to the external tax expert's email address inquiring about the question.

[0038] Furthermore, the LLM 71 refers to the business support system function list (see Figure 4) to identify the tax LLM 741 as a business support system that can execute action 2, and selects the API 741 of the LLM 741 as the destination of the request. The LLM 741 creates a sentence expressed in natural language to inquire of the tax LLM 741 as the content of the request, and sends the request to the API 742 of the tax LLM 741. The tax LLM 741 executes the request and outputs an answer to the question about taxation.

[0039] In the above example, the email sending / receiving system 731 sends a request to the API 732 and the tax LLM 741 sends a request to the API 742 simultaneously, but the process of inquiring by email to a tax expert may be performed only if an appropriate response is not obtained from the tax LLM 741.

[0040] According to this embodiment, by simply inputting a question about tax matters into the LLM in natural language, a plurality of appropriate requests that realize the contents of the output instructions can be executed. [Example]

[0041] Figure 8 is a diagram showing the configuration of a business information processing system according to Example 4. When image information of an invoice is uploaded, the invoice image management system 831 has the function of applying AI-OCR to the invoice image to organize, extract, and save the text data recorded in the invoice image. The payment processing system 833 has the function of creating payment information and sending the created payment information to a financial institution.

[0042] The invoice image management system 831 is pre-programmed to send an output instruction to the LLM81 stating, when invoice image information is uploaded and AI-OCR is applied to the invoice image information, that if it is determined that payment is necessary, "If the request for payment has been approved, please request payment from the financial institution by the invoice deadline."

[0043] When the above output instruction is input to LLM81, action 3 "Confirm whether the request for payment has been approved" is specified as the action for executing the output instruction.

[0044] The LLM 81 refers to the business support system function list (see Figure 4) and identifies the approval system 832 as a business support system that can execute action 3. It then creates a request corresponding to action 3 in a format that can be processed by the API 842 of the approval system 832, and sends the created request to the API 842 of the approval system 832. The approval system 832 executes the request, searches its database to see if there is an approved approval request corresponding to the payment information, notifies the user of this fact if there is an approved approval request, and sends this fact to the API 82 of the LLM 81.

[0045] Next, based on the information that there is an approved request form, the LLM 81 identifies action 4 "If the request form for the payment is approved, execute payment processing" as the next action. The LLM 81 then references the business support system function list (see Figure 4) and identifies the payment processing system 833 as a business support system that can execute action 4. The LLM 81 then creates a request corresponding to action 4 in a format that can be processed by the API 843 of the payment processing system 833, and sends the request to the API 843 of the payment processing system 833. The payment processing system 833 executes the request, creates payment request data, and sends the payment request data to the financial institution that the company uses for payments by the payment deadline.

[0046] In this embodiment, a request is sent to the approval system 832 to inquire whether there is an approved approval form for the transaction, but a request may also be sent to the ordering system to check whether there is a purchase order for the transaction.

[0047] If bill payment work is done manually, it takes a lot of work and there is a risk of missed payments, but according to this embodiment, there is no burden on the user and payments can be made reliably within the payment deadline.

[0048] In addition, output instructions for LLM can be written in natural language, so new output instructions can be easily added according to business needs, even without programming knowledge.

[0049] The embodiments of the present invention are not limited to the above-described embodiments and examples, and various modifications of the above-described embodiments and examples can be made within the scope of the claims. [Explanation of symbols]

[0050] 11 Language Models 12 API (LLM) 13 Business Support System 14 API 20 Computer 21 Memory section 22 Arithmetic section 23 Input section 24 Output section 25 Communications Department 4 Business support system function list 51 LLM 52 API (LLM) 531 Approval System 541, 542 API 531 Approval System 532 Ordering System 61 LLM 62 API (LLM) 631 Accounting System 632 Invoice Image Database 641, 642 API 71 LLM 72 API (LLM) 731 Email sending and receiving system 732, 742 API 81 LLM 82 API (LLM) 831 Invoice Image Management System 832 Approval System 833 Payment Processing System

Claims

1. A language model that has been machine-learned to output an answer when an output instruction is input; A business support system that supports a plurality of specific business operations; An API that is an interface of the business support system; a business support system function list in which functions of the business support system are recorded, a business information processing system in which, when an output instruction is input to the language model, the language model selects the business support system required for output in accordance with the output instruction by referring to the business support system function list, and transmits a request to execute the output instruction to an API of the selected business support system, The business support system includes an approval system and an ordering system, In the approval system, when the approval request is approved, the language model extracts purchase information included in the approval request, and inputs the extracted purchase information into an API, which is an interface of an ordering system, and the ordering system executes a purchase process corresponding to the purchase information input into the API. Business information processing system.

2. A language model that has been machine-learned to output an answer when an output instruction is input; A business support system that supports a plurality of specific business operations; An API that is an interface of the business support system; a business support system function list in which functions of the business support system are recorded, a business information processing system in which, when an output instruction is input to the language model, the language model selects the business support system required for output in accordance with the output instruction by referring to the business support system function list, and transmits a request to execute the output instruction to an API of the selected business support system, the business support system includes an accounting system, When the language model is instructed to compare input information regarding the voucher with information related to the voucher, the language model sends a request to the API of the accounting system, including an instruction to output the input information regarding the voucher, compares the input information regarding the voucher output by the accounting system in accordance with the request with information related to the voucher, and outputs the comparison result. Business information processing system.

3. When the language model is instructed to compare input information related to the voucher with information recorded in an image of the voucher, the language model sends a request to the API of the accounting system including an instruction to output the input information related to the voucher, compares the input information related to the voucher output by the accounting system in accordance with the request with the information recorded in the image of the voucher, and outputs the comparison result.

3. The business information processing system according to claim 2.

4. A language model that has been machine-learned to output an answer when an output instruction is input; A business support system that supports a plurality of specific business operations; An API that is an interface of the business support system; a business support system function list in which functions of the business support system are recorded, a business information processing system in which, when an output instruction is input to the language model, the language model selects the business support system required for output in accordance with the output instruction by referring to the business support system function list, and transmits a request to execute the output instruction to an API of the selected business support system, The business support system includes an invoice management system, an approval system, and a payment processing system, When an invoice image is uploaded to the invoice management system, the language model sends a first request to an API of the approval system, the first request including an output instruction to search for an approved approval document corresponding to the transaction recorded in the invoice image; When the approval system outputs a response indicating that an approved approval document for the transaction exists in response to the first request, the language model sends a second request to an API of the payment processing system, the second request including an output instruction to perform payment processing corresponding to the transaction. Business information processing system.

5. computer-implemented, an output instruction input step of inputting an output instruction into a machine-learned language model that outputs an answer when an output instruction is input; a request sending step of selecting a business support system from a plurality of business support systems by referring to a business support system function list in which functions of the business support systems are recorded, the business support system supporting a specific business required for output by the language model in accordance with the output instruction, and sending a request to an API of the selected business support system; a request execution step of causing the business support system corresponding to the API to which the request was sent in the request sending step to execute the request and output a response from the business support system; A business information processing method comprising: Furthermore, the business support system includes a request system and an ordering system, In the approval system, when an approval request is approved, the language model extracts purchasing information contained in the approval request, inputs the extracted purchasing information into an API, which is an interface of an ordering system, and the ordering system executes purchasing processing corresponding to the purchasing information input into the API. This is a business information processing method having a purchasing processing execution step.

6. A method implemented by a computer, an output instruction input step of inputting an output instruction into a machine-learned language model that outputs an answer when an output instruction is input; a request sending step of selecting a business support system from a plurality of business support systems by referring to a business support system function list in which functions of the business support systems are recorded, the business support system supporting a specific business required for output by the language model in accordance with the output instruction, and sending a request to an API of the selected business support system; a request execution step of causing the business support system corresponding to the API to which the request was sent in the request sending step to execute the request and output a response from the business support system; A business information processing method comprising: Furthermore, the business support system includes an accounting system, a matching result output step in which, when the language model is instructed to match input information regarding a voucher with information related to the voucher, the language model sends a request to the API of the accounting system, including an instruction to output the input information regarding the voucher, matches the input information regarding the voucher output by the accounting system in accordance with the request with the information related to the voucher, and outputs the matching result.

7. A method implemented by a computer, an output instruction input step of inputting an output instruction into a machine-learned language model that outputs an answer when an output instruction is input; a request sending step of selecting a business support system from a plurality of business support systems by referring to a business support system function list in which functions of the business support systems are recorded, the business support system supporting a specific business required for output by the language model in accordance with the output instruction, and sending a request to an API of the selected business support system; a request execution step of causing the business support system corresponding to the API to which the request was sent in the request sending step to execute the request and output a response from the business support system; A business information processing method comprising: Furthermore, the business support system includes an invoice management system, an approval system, and a payment processing system, a first request sending step in which, when an invoice image is uploaded to the invoice management system, the language model sends to an API of the approval system a first request including an output instruction to search for an approved approval document corresponding to the transaction recorded in the invoice image; a second request sending step in which, when the approval system outputs a response indicating that an approved approval document for the transaction exists in response to the first request, the language model sends a second request to an API of the payment processing system, the second request including an output instruction to perform payment processing corresponding to the transaction; A business information processing method comprising the steps of:

8. On the computer, an output instruction input step of inputting an output instruction into a machine-learned language model that outputs an answer when an output instruction is input; a request sending step of selecting a business support system from a plurality of business support systems by referring to a business support system function list in which functions of the business support systems are recorded, the business support system supporting a specific business required for output by the language model in accordance with the output instruction, and sending a request to an API of the selected business support system; a request execution step of causing a business support system corresponding to the API to which the request was sent in the request sending step to execute the request and output a response from the business support system; A business information processing program for executing Furthermore, the business support system includes a request system and an ordering system, In the approval request system, when the approval request is approved, the language model extracts purchase information included in the approval request, and inputs the extracted purchase information into an API, which is an interface of an ordering system, and the ordering system executes a purchase process corresponding to the purchase information input into the API. A business information processing program for execution.

9. A computer comprising: an output instruction input step of inputting an output instruction into a machine-learned language model that outputs an answer when an output instruction is input; a request sending step of selecting a business support system from a plurality of business support systems by referring to a business support system function list in which functions of the business support systems are recorded, the business support system supporting a specific business required for output by the language model in accordance with the output instruction, and sending a request to an API of the selected business support system; a request execution step of causing a business support system corresponding to the API to which the request was sent in the request sending step to execute the request and output a response from the business support system; A business information processing program for executing Furthermore, the business support system includes an accounting system, A business information processing program for executing a matching result output step in which, when the language model is instructed to match input information regarding a voucher with information related to the voucher, the language model sends a request to the API of the accounting system including an instruction to output the input information regarding the voucher, matches the input information regarding the voucher output by the accounting system in accordance with the request with information related to the voucher, and outputs the matching result.

10. A computer comprising: an output instruction input step of inputting an output instruction into a machine-learned language model that outputs an answer when an output instruction is input; a request sending step of selecting a business support system from a plurality of business support systems by referring to a business support system function list in which functions of the business support systems are recorded, the business support system supporting a specific business required for output by the language model in accordance with the output instruction, and sending a request to an API of the selected business support system; a request execution step of causing a business support system corresponding to the API to which the request was sent in the request sending step to execute the request and output a response from the business support system; A business information processing program for executing Furthermore, the business support system includes an invoice management system, an approval system, and a payment processing system, a first request sending step in which, when an invoice image is uploaded to the invoice management system, the language model sends to an API of the approval system a first request including an output instruction to search for an approved approval document corresponding to the transaction recorded in the invoice image; a second request sending step in which, when the approval system outputs a response indicating that an approved approval document for the transaction exists in response to the first request, the language model sends a second request to an API of the payment processing system, the second request including an output instruction to perform payment processing corresponding to the transaction; A business information processing program for executing the above.

Images