CN-121979496-A - Service development framework for realizing AI Agent based on AI, RPA and LLM

CN121979496ACN 121979496 ACN121979496 ACN 121979496ACN-121979496-A

Abstract

The application provides a service development framework for realizing AI Agent based on AI, RPA and LLM, which comprises a basic framework layer, a first service application layer and a second service application layer, wherein the basic framework layer is used for providing various components for the service application layer, the various components are packaged with multiplexing technical components required by various service development, the first service application layer is used for packaging general service application which can be multiplexed through service scene precipitation so as to directly call the general service application for service development, and the second service application layer is used for calling target components and/or general service application required by the service scene in the basic framework layer according to the service scene developed by a user to form customized service application for service development. Therefore, each layer of responsibility of the service development framework for realizing the AI Agent based on the AI, the RPA and the LLM is single, independent expansion and maintenance are convenient, and the multi-service line development efficiency can be remarkably improved on the premise of ensuring the safety and the performance through the layered framework.

Inventors

CAO XI
WU YUANYUAN
ZHANG JINMING
LUO LIANG
WANG GUANCHUN
HU YICHUAN

Assignees

来也科技(北京)有限公司
北京来也网络科技有限公司

Dates

Publication Date: 20260505
Application Date: 20251226

Claims (10)

1. The utility model provides a business development frame based on artificial intelligence AI, robot flow automation RPA and big language model LLM realize Agent AI Agent which characterized in that, business development frame includes basic frame layer, first business application layer and second business application layer, wherein: the base frame layer is used for providing various components for the first service application layer and/or the second service application layer, and various components are packaged with multiplexing technical components required by various service development; The first service application layer is used for packaging the general service application which can be reused through service scene precipitation so as to directly call the general service application for service development; And the second service application layer is used for calling target components and/or general service applications required by the service scene in the basic framework layer according to the service scene developed by the user to form customized service applications so as to develop the service.
2. The business development framework of claim 1 wherein the reusability technique component comprises a generic component, a support component, and a base component, wherein: the universal component comprises an application and server base component, a communication and interface component, a security and monitoring component and a client support component; The support component comprises a data storage and access component, an event processing component, a cache component, a lock mechanism component and a service registration component; The base component comprises a dependency injection component, a log record component, an index collection component, a plug-in mechanism component, a task scheduling component and an operation component.
3. The business development framework of claim 2 wherein: the application and server base component encapsulates the functions of the universal service application/customized service application start initialization, server-side network management and server configuration management; the communication and interface component encapsulates remote procedure call functions among services, basic API model specification interface formats, two-way communication channels and command line interfaces; the security and monitoring component encapsulates security functions, health check and server index collection, request context management functions in request processing, and interceptors; the client support component encapsulates relevant functions of the client and is used for supporting interaction between the client and the server.
4. The business development framework of claim 2 wherein: The support component comprises a data storage and access component, an object relation mapping function for encapsulating database operation, a storage management function for supporting various storage rear ends and a setting management function for configuring user service development system parameters; the event processing component packages a distributed event system supporting the release and subscription of the event; the cache component encapsulates the function of caching access data; the lock mechanism component encapsulates a lock mechanism which controls concurrent access to maintain data consistency; the service registration component encapsulates a registration center for registration and discovery of services, while integrating third party services supporting extended functionality.
5. The business development framework of claim 2 wherein: the dependent injection component encapsulates a dependent injection framework; The log recording component packages a log recording mechanism for recording the running log; The index collection component packages an index collection mechanism for collecting performance and business indexes; The plug-in mechanism component encapsulates a plug-in mechanism that allows system functionality to be extended by a plug-in; the task scheduling component packages a scheduler responsible for scheduling and managing the timing tasks; The operation component encapsulates a configuration management mechanism responsible for processing configuration files, a coroutine support mechanism supporting concurrent processing capacity, an encryption and decryption mechanism guaranteeing data security, an event processing mechanism realizing event-driven programming and a multi-language support mechanism used in different regions.
6. The business development framework of claim 1 wherein the generic business application comprises a tenant management module and an organization management module, wherein: The tenant management module comprises a tenant life cycle management unit, a tenant user management unit, a software authorization management unit and a user authentication unit; The organization management module comprises an organization user unit, a right management unit, an organization information management unit, a notification and message service unit, a single sign-on support unit, a device manufacturer custom support unit, an organization user management unit and an open interface integration unit.
7. The business development framework of claim 6 wherein: the tenant life cycle management unit covers the operations of creating, modifying and deleting the tenant; the tenant user management unit is used for managing users under tenants; the software authorization management unit is used for verifying the use authority of the software; And the user authentication unit is used for carrying out identity authentication on the tenant user.
8. The business development framework of claim 6 wherein: the organization user unit is used for carrying out identity authentication on users in the organization; The authority management unit is used for defining and managing the authority of users in the organization; The organization information management unit is used for creating, modifying and deleting the organization; the notification and message service unit is used for sending various notification messages to users in an organization; the single sign-on support unit is used for organizing single sign-on of users among different service development systems; the device manufacturer customization support unit is used for providing customization support for the device manufacturer; the organization user management unit is used for managing users in the organization; The open interface integration unit is used for providing an open API for integrating the third party system and the user service development system.
9. The business development framework of claim 1 wherein the customized business application encapsulates functional functions corresponding to a plurality of business scenarios, the functional functions being used to invoke target components required for each business scenario.
10. The business development framework of claim 9 wherein: The customized service application is used for determining service development logic according to a customized service scene, and calling a corresponding target component for development according to the service development logic through a function required by the customized service scene so as to obtain a service development result.

Description

Service development framework for realizing AI Agent based on AI, RPA and LLM Technical Field The application relates to the fields of artificial intelligence (ARTIFICIAL INTELLIGENCE, al for short), robot process automation (Robotic Process Automation, RPA for short) and intelligent Agent (ARTIFICIAL INTELLIGENCE AGENT, AI Agent for short), in particular to a business development framework for realizing AI Agent based on AI, RPA and LLM. Background Robot process automation (Robotic Process Automation, RPA for short) is to simulate the operation of a human on a computer through specific "robot software" and automatically execute process tasks according to rules. Artificial intelligence (ARTIFICIAL INTELLIGENCE, AI for short) is a piece of technical science that studies, develops theories, methods, techniques and application systems for simulating, extending and expanding human intelligence. The large language model (Large Language Model, LLM for short) is a model which is trained through massive texts, can recognize human language, execute language class tasks and has a large number of parameters. The Agent (ARTIFICIAL INTELLIGENCE AGENT, AI Agent) is able to sense the environment, make decisions, and perform actions. The distinction from traditional artificial intelligence is the ability to independently think and act, enabling tools to be invoked to accomplish a given goal. AI Agent is based on LLM as its core computational engine, enabling it to talk, perform tasks, reason and exhibit a degree of autonomy. The system has the capability of autonomously understanding, sensing, planning, memorizing and using tools, and can automatically execute complex tasks. The intelligent Agent AI Agent integrated by the LLM driving various AI capabilities can interact with the staff in natural language, understand instructions and demands of the staff and feed back and reply, master field knowledge related to business to complete complex professional work, disassemble the complex task into a plurality of executable tasks and use data and tools to complete the task, and can also cooperate with the staff to complete complex work, so that the digital staff can realize crossing from automation to intellectualization, complete work more efficiently by the staff, and comprehensively realize man-machine cooperation. In the related art, multi-product line parallel service development has become a normal state. Because of the functional positioning, performance requirements and skill differences of development teams, different service lines (such as e-commerce, finance, social contact and the like) often construct independent systems based on different development languages and open source technical stacks, however, the different teams independently realize general functions such as authentication, journaling, monitoring, configuration management and the like, so that code redundancy and maintenance cost are greatly increased, unified interface specifications, data formats and error handling mechanisms are also lacking, a large number of adaptation layers are needed for development when the systems are integrated, and different tool chain supports are needed for deployment, monitoring and fault detection under a multi-language environment, so that the operation and maintenance team burden is increased. Disclosure of Invention The embodiment of the application provides a service development framework for realizing AI Agent based on AI, RPA and LLM, which solves the problems existing in the related technology, and the technical scheme is as follows: The embodiment of the application provides a service development framework for realizing AI agents based on AI, RPA and LLM, which comprises a base framework layer, a first service application layer and a second service application layer, wherein the base framework layer is used for providing various components for the first service application layer and/or the second service application layer, the various components are packaged with multiplexing technical components required by various service development, the first service application layer is used for packaging general service applications which can be multiplexed through service scene precipitation so as to directly call the general service applications for service development, and the second service application layer is used for calling target components and/or general service applications required by service scenes in the base framework layer according to service scenes developed by users so as to form customized service applications for service development. In one embodiment, the multiplexing technology component comprises a general component, a supporting component and a basic component, wherein the general component comprises an application and server basic component, a communication and interface component, a security and monitoring component and a client supporting component, the supporting component comprises a data storag