Information processing system, information processing method, and program

Abstract

To provide an information processing system, an information processing method, and a program for supporting the use and deployment of AI agents. [Solution] An information processing system comprising: an agent information storage unit that stores information about AI agents introduced in each organization; a user information storage unit that stores personnel information for each user; a metadata assignment unit that assigns information indicating the organization from which the AI ​​agent was introduced and information indicating operational constraints to the AI ​​agent as metadata that can be shared between organizations; and an access control unit that identifies AI agents that can be accessed according to the personnel information of a user from among AI agents whose original organization is a different organization from the organization to which the user belongs, based on the operational constraints set in the metadata of each AI agent.

Description

[Technical Field]

[0001] The present disclosure relates to an information processing system, an information processing method, and a program for managing AI agents operated within an organization. [Background technology]

[0002] In recent years, various industries have been trying to introduce AI agents capable of natural language processing in order to improve business efficiency, etc. For example, Patent Document 1 discloses a system that uses an AI agent to generate product improvements based on information posted by users. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2025-042762 Summary of the Invention [Problem to be solved by the invention]

[0004] One of the challenges in promoting the use of AI agents in business operations in organizations such as companies is that the AI ​​agents tend to be used only by specific organizations or specific individuals, such as the development department or the department that initially introduced them, resulting in personalization and slow progress in the deployment of AI agents. Another challenge is the lack of objective information to determine the effectiveness and sharability of AI agents, making it difficult to evaluate the appropriateness of introducing them.

[0005] One of the objectives of the exemplary embodiments of the present disclosure is to provide an information processing system, an information processing method, and a program for supporting the deployment of AI agents. [Means for solving the problem]

[0006] An information processing system according to one aspect of the present disclosure includes: an agent information storage unit that stores information about AI agents introduced in each organization; a user information storage unit that stores personnel information of each user; a metadata assigning unit that assigns information indicating an organization from which the AI ​​agent was introduced and information indicating operational constraints of the AI ​​agent to the AI ​​agent as metadata that can be shared between organizations; The system is equipped with an access control unit that identifies AI agents that can be accessed according to the personnel information of a user from among AI agents whose source organization is a different organization from the organization to which the user belongs, based on the operational constraints set in the metadata of each of the AI ​​agents.

[0007] The information processing system has the above-mentioned features, allowing users to recognize AI agents that can be accessed based on the user's own personnel information from among the AI ​​agents used in other organizations, thereby supporting the cross-organizational deployment of AI agents.

[0008] Other problems and solutions disclosed in the present application will become apparent from the embodiments and drawings of the present disclosure. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating an example of the configuration of an information processing system according to an embodiment of the present disclosure. [Figure 2] FIG. 2 is a block diagram illustrating an example of a hardware configuration of the information processing system shown in FIG. [Figure 3] FIG. 3 is a block diagram illustrating an example of the software configuration of the information processing system shown in FIG. [Figure 4] FIG. 4 is a schematic diagram showing an example of a user interface presenting accessible AI agents. [Figure 5] FIG. 5 is a schematic diagram showing an example of a user interface that presents the operational performance of an AI agent. [Figure 6] FIG. 6 is a flowchart showing an example of information processing executed in the information processing system shown in FIG. DETAILED DESCRIPTION OF THE INVENTION

[0010] An information processing system according to an embodiment of the present disclosure will be described below with reference to the drawings. In the accompanying drawings, identical or similar elements are designated by identical or similar reference symbols and names, and duplicate descriptions of identical or similar elements may be omitted in the description of the embodiment. Note that the contents shown in the drawings are merely examples for explaining the present embodiment and are merely schematic examples for ease of explanation of the present embodiment. The contents of the drawings may be modified or changed within the scope of no technical problem.

[0011] <System Overview> The information processing system of this embodiment is a system for managing AI agents operated by a specific business entity and supporting the deployment of AI agents, such as sharing AI agents across organizations belonging to the business entity.

[0012] Here, "business entity" refers to an organizational form or legal framework for carrying out a specified business activity. Business entities that can operate this system are not limited to single corporations, but also include research institutions, non-profit corporations (school corporations, medical corporations, etc.), organizations without legal status, public organizations, public interest corporations, and associations consisting of multiple organizations or companies, such as business associations. A business entity is considered to have two or more organizations, each with at least one member.

[0013] In this embodiment, an "organization" refers to any classification unit constituting the business entity described above. For example, if the business entity is composed of multiple companies, each company may be the organizational unit, or a classification unit such as a department within each company may be the organizational unit. If the business entity is a single company, the organization may include business departments such as a sales department, a development department, and a human resources department, or may include organizational units divided according to location, such as branch offices. In addition to business departments, the organization may also include groups defined by predetermined classification criteria according to organizational activities or business involvement types, such as a group for classifying sole proprietors, a group for classifying contractors, and a group for classifying collaborators. In other words, an "organization" can be flexibly defined based on predetermined criteria such as department affiliation, job classification, business domain, contract type, location, or collaboration scheme.

[0014] An "AI agent" refers to an artificial intelligence system that autonomously executes predetermined tasks on behalf of a user. An AI agent's basic function is to execute information processing related to a predetermined business task based on user input instructions, predefined settings, and the like, and output the processing results. The specific functions implemented by an AI agent are not particularly limited, and may include one or more functions such as natural language processing, image analysis, data aggregation, predictive analysis, decision-making support, document creation, automatic response, conversational interface, and workflow automation. This embodiment assumes that multiple AI agents are operated within a business entity. For example, AI agents may be used for one or more business tasks, such as automatically creating sales proposals, automatically responding to inquiries from customers, supporting recruitment selection, and supporting the review of development code. Such AI agents may be developed internally by one of the organizations within the business entity, or provided to one of the organizations within the business entity by an external vendor.

[0015] The implementation form of the AI ​​agent used by the business entity is not particularly limited. The AI ​​agent may be incorporated into the business entity's internal system and operated so as to operate within the business entity's internal network, such as an in-house network. Alternatively, the AI ​​agent may be implemented in an external system, such as a cloud platform provided by an external cloud provider, and used using a system integration method such as API integration. In other words, the AI ​​agent can be managed by the system as long as it is available via a network, whether inside or outside the business entity that implements the system.

[0016] In this embodiment, the information processing system will be described in detail using an example in which a single corporate entity is a user entity of the system and manages AI agents operated in each department of the corporate entity. However, this example is merely an example, and use cases to which the system can be applied are not limited to this example.

[0017] When AI agents are introduced or customized independently within each organization within a business entity, the AI ​​agents can become personalized or fragmented within a specific organization or individual, preventing cross-departmental deployment of even AI agents that could contribute to the overall business. Furthermore, if the data assets utilized by the AI ​​agents contain personal and / or confidential information, it may be difficult to deploy the AI ​​agents across departments. In such cases, verification of the sharability and effectiveness of the AI ​​agents is necessary for their horizontal deployment. However, conventional AI asset management systems do not clearly disclose information about the benefits and scope of application of the AI ​​agents, making it difficult to make cross-departmental deployment decisions.

[0018] The information processing system of this embodiment has a function for assigning attribute information to each AI agent as metadata, including the organization from which the agent was introduced, the business for which it is to be used, and operational constraints. It also has a function for identifying AI agents that a user can access from among those introduced by other organizations based on the operational constraints set in the metadata and personnel information managed for each user (affiliated organization, position, etc.). This enables an organizational and objective determination of whether or not to share AI agents, which tend to be operated in a personal or closed manner, and supports the safe and effective use and deployment of AI agents across organizations. Details of this system are described below using examples shown in the drawings.

[0019] <System configuration> As shown in Fig. 1, the information processing system of this embodiment includes a management server 1 and one or more user terminals 2. The management server 1 and the user terminals 2 are connected to each other so that they can communicate with each other via a network NW. In this embodiment, the network NW is primarily assumed to be the Internet, but the network NW is not limited to the Internet and may be constructed using, for example, a public telephone line network, a mobile phone line network, a wireless communication network, Ethernet (registered trademark), or the like. Note that the configuration shown in the figure is an example and is not limited to this.

[0020] <Administration Server 1> The management server 1 is an information processing device that executes various information processing related to the management of AI agents operated by a business entity. The management server 1 may be configured in an on-premise form using a general-purpose computer such as a workstation or personal computer, or may be logically realized by cloud computing. FIG. 2 is a block diagram illustrating an example of the hardware configuration of the management server 1. Note that the illustrated configuration is an example, and the management server 1 may have other configurations. The management server 1 includes at least a processor 10, a memory 11, a storage 12, a transmission / reception unit 13, an input / output unit 14, etc., which are electrically connected to each other via a bus 16.

[0021] The processor 10 is a computing device that controls the overall operation of the management server 1, controls the transmission and reception of data between each element, and performs information processing necessary for application execution and authentication processing. For example, the processor 10 is a CPU (Central Processing Unit) and / or a GPU (Graphics Processing Unit). Each function (means) of the management server 1 is realized by the processor 10 executing a program or the like stored in the storage 12 and deployed in the memory 11.

[0022] The memory 11 includes a main memory configured with a volatile storage device such as a DRAM (Dynamic Random Access Memory), and an auxiliary memory configured with a non-volatile storage device such as a flash memory, an HDD (Hard Disc Drive), etc. The memory 11 is used as a work area for the processor 10, and also stores a BIOS (Basic Input / Output System) that is executed when the management server 1 starts up, various setting information, etc.

[0023] The storage 12 stores various programs such as application programs, and in particular stores programs for executing the various functions of the present system. A database storing data used for each process may also be constructed in the storage 12. For example, a storage unit 120 (described later) is realized as part of the storage area of ​​the memory 11 and / or the storage 12.

[0024] The transmitting / receiving unit 13 is a communication interface that enables the management server 1 to communicate with various information processing terminals such as the user terminal 2 via a communication network. The transmitting / receiving unit 13 may further include a short-range communication interface such as Bluetooth (registered trademark) and BLE (Bluetooth Low Energy) and / or a USB (Universal Serial Bus) terminal.

[0025] The input / output unit 14 is an information input device such as a keyboard, a mouse, etc., and an output device such as a display, etc. The input / output unit 14 may include a touch panel or the like that has both functions of inputting and outputting information, and may also include a printer, a speaker, etc. as output devices.

[0026] A bus 16 is commonly connected to the above elements and transmits, for example, address signals, data signals and various control signals.

[0027] <User device 2> The user terminal 2 is an information processing terminal carried by a member of a business entity (a member refers to a person who belongs to the business entity, regardless of his or her title or position). The user terminal 2 may be, for example, a mobile terminal such as a smartphone or tablet terminal, or a general-purpose computer such as a workstation or personal computer. The user terminal 2 also includes a processor 20, memory 21, storage 22, a transmitter / receiver 23, an input / output unit 24, etc., which are electrically connected to one another via a bus 26. Each element in the hardware configuration of the user terminal 2 can be configured in the same way as the management server 1 shown in FIG. 2, and detailed description of each element of the user terminal 2 will be omitted.

[0028] <Management Server 1 Functions (Software Configuration)> 3 is a block diagram illustrating an example of functions (software configuration) implemented in the information processing system. The management server 1 may include, for example, a setting processing unit 101, a metadata assignment unit 102, an access control unit 103, an expansion candidate output unit 104, a usage application acceptance unit 105, and an aggregation unit 106 as functions realized by the processor 10 executing a program. These functional units are illustrated as functions executed by the processor 10 of the management server 1, but may also be configured to be executed by a processor of another information processing device, such as a user terminal 2, instead of the management server 1. The storage unit 120 of the management server 1 may include various databases, such as an agent information storage unit 121 and a user information storage unit 122.

[0029] The agent information storage unit 121 stores information about the AI ​​agents introduced in each organization of the business entity. Specifically, the agent information storage unit 121 may store basic information about the AI ​​agent, configuration information, information indicating the operational performance of the AI ​​agent (hereinafter referred to as operational performance information), and the like, linked to identification information for uniquely identifying the AI ​​agent.

[0030] For example, the basic information of an AI agent may include information indicating the organization that introduced the AI ​​agent (hereinafter, sometimes referred to as "originating organization information") and information indicating the target business to be performed by the AI ​​agent (hereinafter, sometimes referred to as "target business information"). The "originating organization" here refers to the organization within a business entity that developed or initially introduced the AI ​​agent, and is the organization responsible for management at the time of introduction or the base organization. Furthermore, the target business is defined as the range of tasks to be performed by the AI ​​agent when the AI ​​agent is introduced. The target business information may include business domain names such as "sales proposal creation," "customer support," "personnel selection support," and "financial analysis," more specific task names, text data outlining the task content, and the like. In addition to the above, the basic information may also include historical information such as the agent's name, version information, operation start date, and revision date, as well as information indicating the developer, manager, and technology used (e.g., type of LLM, fine-tuning method, etc.). In addition, when an AI agent is provided by an external cloud provider, information about the provider, such as the provider's identification information, and information used to connect to the AI ​​agent (e.g., API endpoint, communication method, etc.) may be stored as basic information in the agent information storage unit 121.

[0031] The AI ​​agent configuration information also includes information indicating operational constraints on the AI ​​agent (hereinafter, sometimes referred to as operational constraint information). The operational constraints on the AI ​​agent refer to policies, rules, and restrictions established by a business entity regarding access, operation, and data usage by the AI ​​agent. Specifically, the operational constraint information may include information indicating the security classification of data used by the AI ​​agent (e.g., the knowledge base on which the agent performs tasks) and may also include source target user scope information indicating the range of users who can use the AI ​​agent as defined by the source organization. The range of accessible users may be defined, for example, by the user's position (e.g., section manager or above), organizational attributes (e.g., only employees affiliated with the headquarters), contract type (e.g., limited to full-time employees), etc. The operational constraint information may also include information indicating the usage conditions for the AI ​​agent as defined by the source organization, such as the source target user scope information. The usage conditions specified by the introducing organization may include, in addition to the user range, restrictions on the usage location, restrictions on the terminals that use the AI ​​agent, restrictions on time zones, conditions related to data usage (restrictions on data category, conditions on data retention period, restrictions on cooperation with external services), limitations on the functions that can be used, operation authority levels, etc. The agent information storage unit 121 may store operational restriction information indicating the usage conditions specified by the introducing organization in association with the introducing organization information.

[0032] Other restrictive conditions include conditions that limit the hours of access, conditions that limit the data that can be input to the AI ​​agent, etc. The content of the operational constraints adopted by the AI ​​agent is not particularly limited. In addition to the above, the configuration information of the AI ​​agent may also include settings that specify the data items that can be shared between organizations among the information about the AI ​​agent, settings that specify the data items to be collected as operational performance, and templates of prompts to be input to the AI ​​agent, and the settings applied to the AI ​​agent are not necessarily limited.

[0033] The operational performance information may include, for example, log information that records a series of interactions from a user's input to the AI ​​agent to the AI ​​agent's response. The log information may be acquired directly when the AI ​​agent is incorporated into an internal system, or may be acquired as an execution record that includes information sent and received via API (input information to the AI ​​agent, output information from the AI ​​agent) and data accompanying the sent and received information (e.g., processing time, whether communication was successful, number of calls, etc.) when an AI agent on an external cloud is used. Furthermore, the operational performance information may include, but is not necessarily limited to, at least one of the following: statistical information indicating the adoption status of the AI ​​agent by users in each organization; statistical information indicating the frequency of use of the AI ​​agent; information indicating the adoption effect expressed, for example, by KPI improvement rate or ROI (Return on Investment, or the ratio of effect to cost); and an evaluation score indicating the user's evaluation of the AI ​​agent.

[0034] Statistical information indicating the implementation status includes, for example, the number of users, the cumulative number of departments that have implemented the system, and the implementation rate (the ratio of the number of users to the number of employees belonging to the business unit, or the ratio of the number of organizations that have implemented the system to the total number of organizations). Statistical information indicating the frequency of use includes, for example, the number of active users in a given period, the average number of uses (per user or per organization), the average usage time (session time), the number of transactions executed in a given period, the usage ratio by business, and peak usage times. The format for expressing the implementation effects can be set according to the target business of each AI agent, and includes, for example, the change in the number of transactions processed before and after implementation (productivity improvement rate), the reduction rate of business processing time (e.g., a 30% reduction), the reduction rate of human error (quality index), the amount or rate of cost reduction (e.g., a reduction of 2 million yen per year), and the amount of increased sales and avoided opportunity losses (indirect effects). Information indicating the evaluation of the AI ​​agent may include, for example, a satisfaction rating (e.g., a 5-point rating), a rating by category such as operability / response speed / accuracy, free-form feedback (qualitative information), a rating based on quantitative indicators such as NPS (Net Promoter Score), repeat rate, etc. Other operational performance information may also include the number of troubles, usage fees, and operational costs.

[0035] In addition to the data described above, the agent information storage unit 121 also stores information that is used or accumulated when an AI agent is operated by a business entity, and the data structure of the agent information storage unit 122 is not necessarily limited.

[0036] The user information storage unit 122 stores information about users who belong to a business entity. Specifically, the user information storage unit 122 stores the user's personnel information for each user, linked to identification information (user ID) that uniquely identifies the user. The personnel information may include, for example, information indicating the organization to which the user belongs within the business entity, information indicating the user's job title, information indicating the job responsibilities, and information related to a contract between the business entity and the user, such as an employment contract. The job title is the user's position (status) within the organization, and the information indicating the job title may include the job type, job title, job grade, job responsibilities (scope of responsibility), scope of authority, etc. The user information storage unit 122 may also store information indicating the usage history of each user's AI agent.

[0037] The configuration processing unit 101 executes processing related to the registration of various configuration information related to the introduction and operation of each AI agent. The "configuration" here includes the configuration adjustments and operational policy definitions implemented by users such as designated developers or administrators in order to introduce AI agents into an organization and deploy them in a state where they can be used for business tasks.

[0038] In response to a user request, the setting processing unit 101 may output a setting screen to the user terminal 2 and accept the user's setting operations for the AI ​​agent via the setting screen. The setting screen may present an input format for accepting designation of various setting items such as policies, rules, and restriction conditions for using the AI ​​agent. In this case, the setting processing unit 101 executes a process of registering setting information, such as the data range to be used as the knowledge base of the AI ​​agent, the data security classification, the range of users who can access the AI ​​agent, the user's usage authority, the location of responsibility, and prompt templates, in the agent information storage unit 121 in response to the user's setting operations accepted via the setting screen. For example, the range of users who can access the AI ​​agent may be set on a rule-based basis, depending on attributes such as the user's department (e.g., legal department, human resources department, etc.) in the business entity, their position (e.g., manager or above), and their contract type (e.g., full-time employee only).

[0039] If the AI ​​agent is developed using an external tool or provided by an external provider, the setting processing unit 101 may acquire various setting information for the AI ​​agent using system collaboration techniques such as API collaboration. The setting processing unit 101 may also record the history of the setting content and a change log.

[0040] The metadata assignment unit 102 executes a process of assigning predetermined information about an AI agent to the AI ​​agent as metadata that can be shared between organizations. "Assigning metadata" refers to the process of associating predetermined information about the target AI agent with the identification information of the AI ​​agent and setting it in a state that can be searched and / or referenced in each organization of the business entity. Metadata assignment is executed for each AI agent.

[0041] Specifically, the metadata assigning unit 102 assigns, to each AI agent, information indicating the organization from which the AI ​​agent was introduced and information indicating operational constraints as metadata that can be shared between organizations. The operational constraint information that can be shared between organizations may include, for example, information indicating the security classification of the data used by the AI ​​agent and information indicating the usage conditions of the AI ​​agent specified by the organization from which the AI ​​agent was introduced. The usage conditions specified by the organization from which the AI ​​agent was introduced may include, for example, source target user range information indicating the range of users who can use the AI ​​agent specified in the organization from which the AI ​​agent was introduced according to their position, information indicating the usage environment conditions (such as the location of use, terminal, and time constraints), information indicating the conditions for data use, and information indicating the restrictions on the function and operation level. The metadata assigning unit 102 may assign, to the AI ​​agent, source organization information of the AI ​​agent and operational constraint information indicating the usage conditions specified by the organization from which the AI ​​agent was introduced, in association with each other. The above-mentioned operational constraints indicating the security classification of the data and the source target user range are used in the process of identifying accessible AI agents, which is performed for each user by the access control unit 103 (described later).

[0042] The metadata assigning unit 102 may also assign to each AI agent, as metadata that can be shared between organizations, information indicating the target business to be performed by the AI ​​agent and information indicating the operational performance of the AI ​​agent. The operational performance information that can be shared between organizations may include, for example, at least one of statistical information indicating the adoption status of the AI ​​agent by users in each organization, statistical information indicating the frequency of use of the AI ​​agent, information indicating the adoption effect expressed as a KPI improvement rate or ROI, and an evaluation score indicating the user's evaluation of the AI ​​agent. The metadata, such as the target business and operational performance, is used in the process of extracting recommended agents by the deployment candidate output unit 104, which will be described later.

[0043] The data structure of the metadata is not necessarily limited to the above example. Some of the information collected by this system may be defined as data items that can be shared between organizations. In this case, the metadata assignment unit 102 may extract information corresponding to the data item definition of the metadata from the information registered in the agent information storage unit 121 for the target AI agent. Then, a process may be performed to set the extracted information as a data attribute that can be shared between organizations.

[0044] Each piece of information to be used as metadata may be collected by manual input operations by a user, or may be collected from an external system by means of API integration, etc. Metadata related to operational performance may be generated by analyzing log information accumulated through the use of an AI agent in accordance with predefined rules.

[0045] The access control unit 103 identifies, for each user, AI agents accessible to the user based on the user's personnel information. Specifically, the access control unit 103 identifies AI agents accessible to the user based on the personnel information of the user from among AI agents whose original organizations are different from the user's organization, based on operational constraints set in the metadata of each AI agent. The term "one user" here refers to the user to be determined by the access control unit 103. For example, the access control unit 103 receives a request from a user to present a list of AI agents or a request to determine whether or not to allow access to an AI agent, and determines whether or not to allow access to each AI agent for the user who issued the request. Note that the access control unit 103 may not only determine whether or not to allow access to AI agents introduced by organizations other than the user's organization, but may also perform processing to determine whether or not to allow access to an AI agent introduced by the user's organization, based on the operational constraints of the AI ​​agent.

[0046] Whether or not access is permitted is determined based on rules that use operational constraint information included in the metadata. For example, the access control unit 103 may compare the operational constraints indicated in the metadata with the user's personnel information to determine whether or not the user's personnel information conforms to the operational constraints of the AI ​​agent. The access control unit 103 may also determine whether or not a user from a specific organization is permitted to access based on whether or not the operational constraints conform to predetermined criteria. The access determination may be based on one requirement, or on a combination of multiple requirements.

[0047] As a specific example, the access control unit 103 may refer to the security classification of the data used by a single AI agent in the operational constraint information indicated in the metadata corresponding to that AI agent, and if the security classification is equal to or greater than a predetermined threshold (for example, if the security classification is "top secret" and data is permitted to be disclosed only to a very small number of personnel, such as the development manager or management), determine that access is not permitted and exclude that single AI agent from the AI ​​agents accessible to the user. Alternatively, the access control unit 103 may refer to the user's job title indicated in the personnel information and the security classification of the data used by that single AI agent, and if the access authority to the data specified by the user's job title does not satisfy the security classification indicated in the operational constraint information, determine that access is not permitted and exclude that single AI agent from the AI ​​agents accessible to the user.

[0048] The access control unit 103 may also use operational constraint information corresponding to the source organization information, particularly information indicating the terms of use specified by the source organization, to compare with personnel information of users belonging to organizations other than the source organization, and identify AI agents accessible to the user based on the comparison results. For example, the access control unit 103 may use source target user range information, specified according to job titles in the source organization, among the operational constraint information, to determine whether access is permitted. In this case, the access control unit 103 may use source target user range information indicated in the metadata of each AI agent to identify the AI ​​agents accessible to the user. For example, if the target user range is set to those with a position of section chief or higher in the source organization, the access control unit 103 may also use the target user range in other organizations to determine whether access is permitted to users of other organizations who hold a position of section chief or higher. Alternatively, if the operational constraint information restricts the organizations that can use the AI ​​agent, access may be determined based on whether the user's organization is one of the permitted organizations. The access determination rules are not necessarily limited to the above examples. The access control unit 103 may determine whether or not to allow access based on the user's usage record of the AI ​​agent, a history of violations of operational rules, and the like.

[0049] In this system, access permission for each AI agent may be determined based on rules as described above, and then an organization's administrator may edit the access permissions of each member. For example, the access control unit 103 may present an organization-specific list (a list showing the access permission settings for each organization's users) to an administrative user who manages a specific organization, indicating the AI ​​agents that each user belonging to the specific organization can access. Then, via the organization-specific list, the administrative user may perform an editing operation to edit the access permission settings for each AI agent for each user, and the access permission settings for each AI agent for each user may be changed in accordance with the editing operation.

[0050] The expansion candidate output unit 104 executes a process of presenting to the user the AI ​​agents that are accessible to the user identified by the access control unit 103, along with metadata related to the AI ​​agents. Specifically, in response to a user request, the expansion candidate output unit 104 may output to the user terminal 2 a display screen (user interface) showing a list of AI agents that have been introduced in the business entity, and present detailed information about the AI ​​agents that the user can use via the list display screen.

[0051] For example, FIG. 4 illustrates an example of a list display screen of AI agents presented to a user. In the example of FIG. 4, the target business of each AI agent, the organization from which it was introduced, the introduction status such as the number of users, the introduction effect such as the KPI improvement rate, and a recommendation flag are displayed for each agent. Note that in addition to presenting accessible AI agents, the list display screen as shown in FIG. 4 may also present AI agents determined to be inaccessible. In this case, each AI agent is displayed in a manner that allows the user to identify whether or not it is accessible (for example, by displaying a flag indicating whether or not it is accessible). The deployment candidate output unit 104 dynamically controls the display mode of the list as shown in FIG. 4 according to the accessibility determination result determined for each user by the access control unit 103. Via the UI output by the deployment candidate output unit 104, the user can identify candidate agents that can be used in their own business from among AI agents operated by other organizations, etc.

[0052] The deployment candidate output unit 104 may not only present AI agents accessible to the user, but may also select from among those AI agents an AI agent that it recommends for use by the user (hereinafter referred to as a recommended agent). Specifically, the deployment candidate output unit 104 may extract recommended agents to be recommended as introduction candidates to a single user from among the AI ​​agents identified by the access control unit 103, based on the target business indicated in the metadata of each AI agent and / or its operational track record. The extraction of recommended agents may be performed on a rule-based basis, in which case the extraction rules are not necessarily limited.

[0053] For example, the expansion candidate output unit 104 may calculate the task compatibility between the tasks indicated in the personnel information of a user and the target tasks indicated in the metadata of each AI agent. Then, recommended agents for the user may be extracted based on the task compatibility. The task compatibility may be calculated based on a comparison table indicating the similarity between tasks, which is prepared in advance, or may be calculated based on the usage history of the AI ​​agent for each task; the calculation method is not particularly limited. Furthermore, the task compatibility may be expressed as a graded index such as "low, medium, high," and the expression format is not particularly limited. The expansion candidate output unit 104 may extract AI agents whose task compatibility is equal to or greater than a predetermined threshold (for example, task compatibility of "medium" or greater) as recommended agents.

[0054] The deployment candidate output unit 104 may also extract recommended agents based on operational performance. For example, the deployment candidate output unit 104 may set priorities for AI agents based on the KPI improvement rate or ROI included in the operational performance indicated in the metadata of each AI agent. The priorities are the order of AI agents recommended to a user, and are determined based on the implementation effect indicated by the KPI improvement rate or ROI. For example, the deployment candidate output unit 104 extracts recommended agents based on the set priorities, such as by setting AI agents with a priority level higher than a predetermined level as recommended agents. The method of extracting recommended agents is not limited to the above examples. For example, recommended agents may be extracted based on the number of users in the entire business entity or the trend in the number of users over a recent predetermined period. Recommended agents may also be extracted based on the number of users by other users in the user's organization or by other users with a similar position to the user, or AI agents with reported successful implementation cases may be extracted as recommended agents.

[0055] After extracting the recommended agents, the expansion candidate output unit 104 may display the recommended agents in a list of AI agents such as that shown in Fig. 4 in a manner that allows them to be identified. For example, in Fig. 4, a flag indicating the recommended agent is attached. Alternatively, the AI ​​agents may be sorted so that the recommended agents are placed at the top of the list, or the recommended agents may be highlighted. Alternatively, the expansion candidate output unit 104 may filter and display only the recommended agents on the UI.

[0056] The deployment candidate output unit 104 may not only display the AI ​​agents in a list format as shown in FIG. 4, but may also output a UI that presents the analysis results of the operational performance in a dashboard format. For example, the deployment candidate output unit 104 may output a heat map by referring to statistical information indicating the implementation status of AI agents by users in each organization, among the operational performance information. In this case, the deployment candidate output unit 104 outputs the implementation status as a heat map color-coded by organization based on the implementation status indicated in the metadata of each AI agent. FIG. 5(a) is an example of a heat map color-coded by organization, in which the implementation status of AI agents in each organization is indicated by different shades of color. Such a heat map allows the deployment status of each AI agent in each organization to be visually and intuitively grasped.

[0057] In addition, a chart plotting the operational performance of an AI agent on multiple axes may be output. For example, the deployment candidate output unit 104 may output a chart plotting two or more of the following on multiple axes: business compatibility with a user's business, security classification, implementation status, frequency of use, implementation effect, and evaluation score for each AI agent. Figure 5(b) shows an example of a radar chart with the evaluation score, implementation status, frequency of use, implementation effect (KPI improvement rate or ROI), security classification, and business compatibility for each AI agent as axes. Such a chart allows the user to intuitively compare and consider the strengths and weaknesses of each agent.

[0058] The method of visualizing the implementation results on the UI is not limited to the above example. The deployment candidate output unit 104 may output, as graphs to be presented on the dashboard, a line graph showing the implementation trend of each AI agent over time, a bar graph showing the usage frequency by user, a histogram showing the distribution of evaluation scores, a bubble chart comparing the implementation effects by target business, etc. The deployment candidate output unit 104 may accept an input operation from the user specifying display conditions, and execute a process of switching the display format of the graph to be presented on the dashboard in accordance with the input operation.

[0059] The usage application receiving unit 105 receives a selection operation from the user of a new AI agent that the user wishes to use, and executes processing related to the usage application to start using the selected AI agent. Here, "usage application" refers to the procedure related to starting use of the target AI agent selected by the user.

[0060] The usage request receiving unit 105 may receive the selection of an AI agent to be requested for use via a UI output by the deployment candidate output unit 104. For example, on the AI ​​agent list display screen shown in FIG. 4, an operation button such as "Apply" may be presented in the display column of each AI agent. The usage request acceptance process may then be executed through a user's input operation on the operation button. Also, an operation area for requesting a usage request to each AI agent may be provided on a dashboard including a graph such as that shown in FIG. 5. In the UI output by the deployment candidate output unit 104, the operation area for accepting a usage request may be controlled so that it is active only for AI agents that the access control unit 103 has determined to be accessible to the user.

[0061] The usage application receiving unit 105 may execute a predetermined approval flow after receiving a usage application for the selected AI agent. For example, a configuration may be adopted in which an approval request for the usage application for the selected AI agent is sent to the administrator of the organization to which the user belongs. Note that even if the AI ​​agent is determined to be accessible by the access control unit 103, a process may be required to determine whether it can be used before the first use. The usage application receiving unit 105 may execute a process to select a plan and mediate a usage contract when starting to use the AI ​​agent.

[0062] The aggregation unit 106 executes a process for generating operational performance information for AI agents. For example, the aggregation unit 106 executes a process for calculating various operational performance data for the AI ​​agent, such as its implementation status, usage frequency, implementation effects, and evaluation score, by collecting statistics and / or analyzing the usage logs of the AI ​​agent by each user. This operational performance data is assigned as metadata for each AI agent by the metadata assignment process described below.

[0063] For example, the aggregation unit 106 may calculate statistical information regarding the implementation status and usage frequency by aggregating the number of calls to the AI ​​agent within a predetermined period, the frequency of use by each user, the number of users who have already implemented the AI ​​agent, and the implementation status by department. Furthermore, the implementation effects, such as the KPI improvement rate and business return on investment (ROI), may be calculated by analyzing performance data acquired from other systems for business management and / or performance data entered by users. Furthermore, the aggregation unit 106 may execute a process of aggregating evaluation scores entered by users after using the AI ​​agent and calculating the statistical values.

[0064] The process of generating the performance information may be automatically executed by the aggregation unit 106, or the user may edit or correct part or all of the information as desired. Information that cannot be obtained by log analysis or the like may be analyzed by an external tool and used, or may be manually registered by the user. The performance information may be updated as appropriate in response to user instructions.

[0065] Furthermore, the performance information may be automatically re-aggregated at a predetermined update interval or upon the occurrence of a predetermined event. The update interval may be a regular schedule, such as daily, weekly, or monthly, and is not limited to this. Examples of predetermined events include, but are not limited to, when the number of uses of a specific AI agent exceeds a predetermined threshold, when the performance rules are revised, or when a new target task is added. An event-triggered method may be employed, in which the re-aggregation of performance information is triggered by the occurrence of such an event. The update process for the performance information may verify the consistency of the acquired log data and performance data received from an external system, and correct missing and / or abnormal values ​​before re-aggregation. By updating the performance information periodically or upon the occurrence of a certain condition, the performance information assigned as metadata is always kept up-to-date and highly reliable, thereby improving the accuracy of the judgment by the access control unit 103 and the presentation by the deployment candidate output unit 104.

[0066] In the above, some or all of the functions of the setting processing unit 101, metadata assignment unit 102, access control unit 103, deployment candidate output unit 104, usage application reception unit 105, and aggregation unit 106 shown as functional units of the management server 1 may be configured to be realized by the processor 20 of the user terminal 2.

[0067] The UI operation reception unit 201 has a function of receiving input operations from the user via various user interfaces (for example, UIs such as those shown in FIGS. 4 and 5 output by the deployment candidate output unit 104) displayed on the user terminal 2. The UI operation reception unit 201 acquires operation signals via various input means provided on the user terminal 2, such as a touch panel, a mouse, a keyboard, or a pointing device, and executes a process of transmitting the operation signals to the management server 1 via the transmission / reception unit 23.

[0068] <Example of information processing method> Next, an example of an information processing method executed by the information processing system of this embodiment will be described with reference to the flowchart illustrated in FIG.

[0069] First, the configuration processing unit 101 of the management server 1 executes a process to register various configuration information for the AI ​​agent to be applied to the source organization (step SQ101). The configuration information may be specified based on input operations by users such as developers or administrators, or may be obtained by means of API integration or other means from an external provider that provides the AI ​​agent. The AI ​​agent configuration process specifies information such as the target business of the AI ​​agent, usage policy, operational constraints, and purpose of introduction, and may also set target user range information specified according to job position and other factors in the source organization.

[0070] Next, the metadata assignment unit 102 executes a process of assigning predetermined information about the AI ​​agent to the AI ​​agent as metadata that can be shared between organizations (step SQ102). Specifically, information indicating the organization from which the AI ​​agent was introduced and information indicating operational constraints of the AI ​​agent (such as the security classification of data used by the AI ​​agent and the target user range of the AI ​​agent) are assigned to each AI agent as metadata. In addition, information indicating the target tasks to be performed by the AI ​​agent and information indicating the operational performance of the AI ​​agent may also be associated as metadata.

[0071] Next, the access control unit 103 executes a process to identify AI agents accessible to each user based on the user's personnel information (step SQ103). In this process, the AI ​​agents accessible to the user are identified from among AI agents whose source organization is different from the user's organization, based on the operational constraints set in the metadata of each AI agent. The access permission determination may be rule-based, using predetermined determination rules based on the operational constraints. For example, the access control unit 103 may determine that the user is not allowed to access the AI ​​agent if the security classification of the data used by the AI ​​agent is equal to or higher than a predetermined threshold, or if the access authority defined according to the user's position does not meet the security classification of the AI ​​agent. In addition, the target user range (settings defining the range of users who can use the AI ​​agent) set in the source organization may be applied to other organizations to determine whether the user is allowed to access the AI ​​agent.

[0072] The deployment candidate output unit 104 then executes a process of extracting recommended agents to be recommended to the user from among the AI ​​agents identified as accessible in step SQ103 (step SQ104). In this recommendation process, recommended agents may be extracted based on the degree of compatibility between the target task indicated in the metadata of each AI agent and the task for which the user is responsible. Alternatively, a priority order may be set indicating the ranking of AI agents to be recommended based on operational performance such as KPI improvement rate and ROI, and recommended agents may be extracted according to that priority order.

[0073] Next, the deployment candidate output unit 104 executes a process of outputting a user interface (UI) for presenting accessible AI agents (step SQ105). This UI is, for example, a list display screen as shown in FIG. 4, and presents metadata including the introduction information, target business, and recommendation flags of each AI agent in a visible format. The deployment candidate output unit 104 may also output a dashboard that visually shows the operational performance, such as a heat map (FIG. 5(a)) that visualizes the introduction status, or a radar chart (FIG. 5(b)) that plots operational performance data on multiple evaluation axes. The UI output by the deployment candidate output unit 104 may also allow a user to apply for new use of an AI agent.

[0074] It should be noted that the flowchart in Fig. 6 is merely an example, and the information processing method of this embodiment is not limited to the example in Fig. 6. Steps may be added, deleted, changed, or their order may be rearranged as appropriate.

[0075] In the information processing system of this embodiment, metadata such as operational constraint information is assigned to AI agents distributed and deployed in each organization within the business, enabling unified management of AI assets within the business. Furthermore, by controlling accessible AI agents based on the operational constraints assigned as metadata and the user's personnel information (affiliated organization, job position, etc.), each organization within the business can safely and efficiently use and deploy AI agents from other organizations based on appropriate information for decision-making.

[0076] In addition, by suggesting recommended agents to users based on the operational performance and / or business suitability of AI agents used in other organizations, it is possible to more effectively promote the use and deployment of AI agents across organizations. Furthermore, by presenting operational performance such as the usage status and introduction effects of each AI agent in a visual UI such as a heat map or radar chart, it becomes easier to understand the uneven distribution of AI agent use within a business entity, which is expected to promote the strategic reallocation of AI resources and the deployment of AI agents.

[0077] The above-described embodiments are merely examples for facilitating understanding of the present disclosure, and are not intended to limit the present disclosure. The present disclosure can be modified or improved without departing from the spirit thereof, and it goes without saying that the present disclosure includes equivalents thereof.

[0078] For example, the access control unit 103 may determine whether to grant access by taking into consideration not only the user's personnel information but also performance information such as the user's work performance status and participation history in recent projects. Furthermore, the display format of the dashboard output by the deployment candidate output unit 104 is not limited to the example shown in FIG. 5 , and the user may freely switch between graph types (e.g., pie chart, stacked bar graph, scatter plot, etc.). Visualization of the operational performance of an AI agent may use a matrix display that allows comparison of the implementation effects in multiple organizations. Furthermore, the application process when a user wishes to use an AI agent may not necessarily require an explicit "application operation," and the initial call operation of the AI ​​agent may be considered as the application. Alternatively, instead of manual approval by an administrator, the application may be automatically approved if predetermined conditions (e.g., compatibility with the target business, evaluation score, operational risk, etc.) are met.

[0079] Furthermore, the series of processes performed by the information processing system described herein may be implemented using software, hardware, or a combination of software and hardware. A computer program for implementing each function of the management server 1 according to this embodiment may be created and installed on a PC or the like. A computer-readable recording medium on which such a computer program is stored may also be provided. Examples of the recording medium include a magnetic disk, an optical disk, a magneto-optical disk, and a flash memory. Furthermore, the computer program may be distributed, for example, via a network, without using a recording medium.

[0080] Furthermore, the effects described herein are merely descriptive or exemplary and are not limiting. In other words, the technology according to the present disclosure may achieve other effects that are apparent to those skilled in the art from the description of this specification, in addition to or in place of the above-described effects.

[0081] The information processing system, information processing method, and program of the present disclosure may have the following configuration. [Item 1] an agent information storage unit that stores information about AI agents introduced in each organization; a user information storage unit that stores personnel information of each user; a metadata assigning unit that assigns information indicating an organization from which the AI ​​agent was introduced and information indicating operational constraints of the AI ​​agent to the AI ​​agent as metadata that can be shared between organizations; An information processing system comprising: an access control unit that identifies AI agents that can be accessed based on the personnel information of a user from among AI agents whose source organization is a different organization from the organization to which the user belongs, based on the operational constraints set in the metadata of each of the AI ​​agents. [Item 2] The information indicating the operational constraints includes information indicating a security classification of data used by the AI ​​agent, The information processing system described in item 1, wherein the access control unit excludes a single AI agent from the AI ​​agents accessible by the single user if the security classification indicated in the metadata corresponding to the single AI agent is above a predetermined threshold, or if the access rights to data specified by the position indicated in the personnel information of the single user do not meet the security classification indicated in the metadata. [Item 3] The information indicating the operational constraints includes source target user range information indicating the range of users who can use the AI ​​agent specified by the source organization, The information processing system of claim 1, wherein the access control unit converts the source target user range information indicated in the metadata of each of the AI ​​agents to the organization to which the one user belongs, thereby identifying AI agents that the one user can access. [Item 4] The information processing system described in any of items 1 to 3, wherein the access control unit presents an organizational list to an administrative user who manages a specified organization, showing the AI ​​agents that can be accessed by each affiliated user belonging to the specified organization, and accepts editing operations by the administrative user to edit the accessibility of each affiliated user to each AI agent via the organizational list. [Item 5] the metadata assigning unit assigns, to the AI ​​agent, source organization information indicating the source organization that introduced the AI ​​agent and operation constraint information indicating the usage conditions stipulated by the source organization in association with each other; The information processing system described in item 1, wherein the access control unit converts the operational constraint information corresponding to the source organization information into personnel information of a user who belongs to an organization other than the source organization and compares it, and identifies AI agents that the user can access based on the comparison results. [Item 6] The information processing system described in Item 1, wherein the metadata assignment unit further assigns to the AI ​​agent, as metadata that can be shared between organizations, information indicating the target task to be performed by the AI ​​agent and information indicating the operational performance of the AI ​​agent. [Item 7] The information processing system described in item 6 further comprises an expansion candidate output unit that presents to the one user the AI ​​agent identified as accessible to the one user along with the metadata of the AI ​​agent. [Item 8] The information processing system described in Item 7, wherein the deployment candidate output unit extracts recommended agents to be recommended as introduction candidates to the one user from among the AI ​​agents identified by the access control unit based on the target business indicated in the metadata of each AI agent and / or the operational track record. [Item 9] The information processing system described in item 8, wherein the deployment candidate output unit calculates the degree of business compatibility between the work responsibilities indicated in the personnel information of the one user and the target work indicated in the metadata of each of the AI ​​agents, and extracts the recommended agent for the one user based on the degree of business compatibility. [Item 10] The information processing system described in item 8, wherein the deployment candidate output unit extracts the recommended agent for the one user by setting a priority of the AI ​​agent based on the KPI improvement rate or ROI included in the operational performance indicated in the metadata of each of the AI ​​agents. [Item 11] The information indicating the operational performance includes statistical information indicating the adoption status of the AI ​​agent by users of each organization, The information processing system described in any one of items 7 to 10, wherein the deployment candidate output unit outputs the introduction status as a heat map color-coded by organization based on the introduction status indicated in the metadata of each of the AI ​​agents. [Item 12] The information indicating the operational constraints includes information indicating a security classification of data used by the AI ​​agent, The information indicating the operational performance includes at least one of statistical information indicating the adoption status of the AI ​​agent by users in each organization, statistical information indicating the frequency of use of the AI ​​agent, information indicating the introduction effect expressed as a KPI improvement rate or ROI, and an evaluation score indicating the evaluation result of the AI ​​agent by users; The information processing system described in any one of items 7 to 10, wherein the deployment candidate output unit outputs a chart plotting two or more of the business suitability of each of the AI ​​agents with the user's work, the security classification, the implementation status, the frequency of use, the implementation effect, and the evaluation score on multiple axes. [Item 13] Storing information about the AI ​​agents deployed in each organization; storing personnel information of each user for each user; assigning information indicating the organization from which the AI ​​agent was introduced and information indicating operational constraints of the AI ​​agent to the AI ​​agent as metadata that can be shared between organizations; An information processing method executed by a computer, which identifies AI agents that can be accessed according to the personnel information of a user from among AI agents whose source organization is a different organization from the organization to which the user belongs, based on the operational constraints set in the metadata of each of the AI ​​agents. [Item 14] Storing information about the AI ​​agents deployed in each organization; storing personnel information of each user for each user; assigning information indicating the organization from which the AI ​​agent was introduced and information indicating operational constraints of the AI ​​agent to the AI ​​agent as metadata that can be shared between organizations; A program for causing a computer to execute the following steps: Identifying an AI agent that can be accessed according to the personnel information of a user from among AI agents whose source organization is a different organization from the organization to which the user belongs, based on the operational constraints set in the metadata of each of the AI ​​agents. [Explanation of symbols]

[0082] 1 Management Server 121 Agent information storage unit 102 Metadata assignment unit 103 Access control section

Claims

1. an agent information storage unit that stores information about AI agents introduced in each organization; a user information storage unit that stores personnel information of each user; a metadata assigning unit that assigns information indicating an organization from which the AI ​​agent was introduced and information indicating operational constraints of the AI ​​agent to the AI ​​agent as metadata that can be shared between organizations; an access control unit that identifies an AI agent that can be accessed according to the personnel information of a user from among AI agents whose introduction source organization is a different organization from the organization to which the user belongs, based on the operational constraints set in the metadata of each of the AI ​​agents; The metadata assignment unit further assigns to the AI ​​agent, as the metadata that can be shared between organizations, information indicating the target business to be performed by the AI ​​agent and information indicating the operational performance of the AI ​​agent. An information processing system.

2. The information indicating the operational constraints includes information indicating a security classification of data used by the AI ​​agent, 2. The information processing system of claim 1, wherein the access control unit excludes a single AI agent from among the AI ​​agents accessible by the single user if the security classification indicated in the metadata corresponding to the single AI agent is equal to or greater than a predetermined threshold, or if the access authority to data defined by the position indicated in the personnel information of the single user does not meet the security classification indicated in the metadata.

3. The information indicating the operational constraints includes introduction source target user range information indicating the range of users who can use the AI ​​agent specified by the introduction source organization, 2. The information processing system according to claim 1, wherein the access control unit diverts the introduction source target user range information indicated in the metadata of each of the AI ​​agents to the organization to which the one user belongs, thereby identifying AI agents that the one user can access.

4. The information processing system according to any one of claims 1 to 3, wherein the access control unit presents an organizational list to an administrative user who manages a specified organization, showing the AI ​​agents that can be accessed by each user belonging to the specified organization, and accepts editing operations by the administrative user to edit the accessibility of each user to each AI agent via the organizational list.

5. The metadata assignment unit assigns, to the AI ​​agent, introduction source organization information indicating an introduction source organization that introduced the AI ​​agent and operation constraint information indicating usage conditions stipulated in the introduction source organization in association with each other, 2. The information processing system according to claim 1, wherein the access control unit converts the operational constraint information corresponding to the source organization information into personnel information of a user who belongs to an organization different from the source organization, and compares the information with the personnel information of a user who belongs to an organization different from the source organization, and identifies an AI agent that the user can access based on the comparison result.

6. The information processing system according to claim 1 , further comprising an expansion candidate output unit that presents the AI ​​agent identified as accessible to the one user to the one user together with the metadata of the AI ​​agent.

7. 7. The information processing system according to claim 6, wherein the deployment candidate output unit extracts recommended agents to be recommended to the one user as candidates for introduction from among the AI ​​agents identified by the access control unit, based on the target business indicated in the metadata of each of the AI ​​agents and / or the operational performance.

8. 8. The information processing system according to claim 7, wherein the deployment candidate output unit calculates a degree of business compatibility between the work assigned to the one user indicated in the personnel information and the target work indicated in the metadata of each of the AI ​​agents, and extracts the recommended agent for the one user based on the degree of business compatibility.

9. 8. The information processing system according to claim 7, wherein the deployment candidate output unit extracts the recommended agent for the one user by setting a priority order for the AI ​​agents based on a KPI improvement rate or ROI included in the operational performance indicated in the metadata of each of the AI ​​agents.

10. The information indicating the operational performance includes statistical information indicating the adoption status of the AI ​​agent by users of each organization, The information processing system according to any one of claims 6 to 9, wherein the deployment candidate output unit outputs the introduction status as a heat map color-coded by organization based on the introduction status indicated in the metadata of each of the AI ​​agents.

11. The information indicating the operational constraints includes information indicating a security classification of data used by the AI ​​agent, The information indicating the operational performance includes at least one of statistical information indicating the introduction status of the AI ​​agent by users of each organization, statistical information indicating the frequency of use of the AI ​​agent, information indicating the introduction effect expressed by a KPI improvement rate or ROI, and an evaluation score indicating the evaluation result of the AI ​​agent by users; The information processing system according to any one of claims 6 to 9, wherein the deployment candidate output unit outputs a chart in which two or more of the suitability of each of the AI ​​agents for the user's work, the security classification, the implementation status, the frequency of use, the implementation effect, and the evaluation score are plotted on multiple axes.

12. storing information about AI agents deployed in each organization; storing personnel information of each user for each user; Assigning information indicating the organization from which the AI ​​agent was introduced, information indicating operational constraints of the AI ​​agent, information indicating the target business to be executed by the AI ​​agent, and information indicating the operational performance of the AI ​​agent to the AI ​​agent as metadata that can be shared between organizations; An information processing method executed by a computer, which identifies an AI agent that can be accessed according to the personnel information of a single user from among AI agents that have been introduced into an organization different from the organization to which the single user belongs, based on the operational constraints set in the metadata of each of the AI ​​agents.

13. storing information about AI agents deployed in each organization; storing personnel information of each user for each user; Assigning information indicating the organization from which the AI ​​agent was introduced, information indicating operational constraints of the AI ​​agent, information indicating the target business to be executed by the AI ​​agent, and information indicating the operational performance of the AI ​​agent to the AI ​​agent as metadata that can be shared between organizations; A program for causing a computer to execute the following steps: Identifying an AI agent that can be accessed according to the personnel information of a single user from among AI agents that have been introduced into an organization different from the organization to which the single user belongs, based on the operational constraints set in the metadata of each of the AI ​​agents.

Images