Intelligent agent building workbench based on multi-intelligent agent framework

Abstract

The invention discloses an agent building workbench based on a multi-agent framework, and relates to the technical field of artificial intelligence. The system comprises an agent building module, a workflow configuration module, a release and integration module and an enterprise WeChat access module. The agent building module is used for quickly creating an agent according to business requirements and configuring parameters, a function module and a knowledge base of the agent at the same time; the workflow configuration module constructs a task allocation mechanism based on an open source multi-agent architecture, and the workflow configuration module is used for supporting configuration of a multi-agent workflow through a graphical interface and a programming interface, and embedding an open source tool and a self-research tool in the workflow to realize modular processing of tasks; by constructing the Roles intelligent agent construction workbench, the problem that an existing intelligent agent development tool is insufficient in flexibility and expansibility is solved, an efficient intelligent agent construction and management platform is provided for enterprise users, intelligence and automation of the service process are achieved, and the method has wide market application prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence, and specifically to an agent building workbench based on a multi-agent framework. Background Art

[0002] With the rapid development of artificial intelligence technology, the demand for agents in enterprise daily operations is increasing;

[0003] However, most of the existing agent development and management tools are for specific functions, lacking flexibility and scalability, and it is difficult to meet complex business requirements. In addition, enterprise users often need to integrate multiple tools and systems, making the process of building agents and constructing workflows complex and inefficient;

[0004] Currently, there is a lack of a general workbench in the market that can quickly and efficiently support the combination of multiple agents and can flexibly embed open-source and self-developed tools, so as to achieve intelligent task allocation and efficient execution;

[0005] In view of the above problems, the inventor proposes an agent building workbench based on a multi-agent framework to solve the above problems. Summary of the Invention

[0006] In order to solve the problems of high complexity, insufficient flexibility, and poor integration in agent building in the prior art; the purpose of the present invention is to provide an agent building workbench based on a multi-agent framework.

[0007] To solve the above technical problems, the present invention adopts the following technical solution: An agent building workbench based on a multi-agent framework, the workbench includes an agent building module, a workflow configuration module, a publishing and integration module, and an enterprise WeChat access module;

[0008] The agent building module is used to quickly create agents according to business requirements, and at the same time configure the parameters, function modules, and knowledge bases of the agents;

[0009] The workflow configuration module constructs a task allocation mechanism based on the open-source multi-agents architecture. The workflow configuration module is used to support the configuration of multi-agent workflows through a graphical interface and programming interfaces, and embed open-source tools and self-developed tools in the workflow to achieve modular processing of tasks;

[0010] The publishing and integration module is used to publish the configured agents and workflows to the cloud for other users to use. The publishing and integration module optimizes the agents and workflows through the ChainGen framework, and at the same time supports embedding into other agents and workflows to achieve seamless cross-platform integration;

[0011] The WeCom access module is used to support the access of the agent to WeCom, realize the collection and analysis of WeCom chat records, and automatically incorporate the chat records into the agent knowledge base to improve the knowledge management ability.

[0012] Preferably, the workbench supports the combination of multiple agents and can embed open-source tools and self-developed tools.

[0013] Preferably, the workbench collects chat records through the WeCom access module and saves them to the agent knowledge base.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0015] 1. In the present invention, by building a workbench for the Kaiwei Roles agent, the problems of insufficient flexibility and scalability of existing agent development tools are solved, providing an efficient agent building and management platform for enterprise users, realizing the intelligentization and automation of business processes, and having broad market application prospects;

[0016] 2. In the present invention, through the cooperation of the agent building module, workflow configuration module, publishing and integration module, and WeCom access module, the development efficiency of the agent is improved, the technical threshold is reduced, workflow support for multiple agents is provided, the requirements of complex business scenarios are met, multiple open-source tools and self-developed tools are compatible, the scalability of the platform is enhanced, WeCom integration is supported, and the enterprise internal knowledge management and business process automation capabilities are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0018] Figure 1 It is the Kaiwei ChainGen agent architecture for the embodiments of the present invention.

[0019] Figure 2 It is the data cleaning and ChatBI architecture for the embodiments of the present invention.

[0020] Figure 3 It is the AIGC agent architecture for the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0022] Embodiment: As Figures 1-3 shown, the present invention provides an agent building workbench based on a multi-agent framework. The workbench includes an agent building module, a workflow configuration module, a publishing and integration module, and a WeCom access module;

[0023] The agent building module is used to quickly create an agent according to business requirements, and at the same time configure the parameters, functional modules, and knowledge base of the agent. Through the workbench interface, the business requirements and configuration parameters can be input by the agent building module. The agent building module generates an agent framework according to the user input and supports user adjustment. The user uploads or associates the knowledge base, and the agent automatically completes knowledge learning and adaptation;

[0024] The workflow configuration module constructs a task assignment mechanism based on the open-source multi-agents architecture. The workflow configuration module is used to support the configuration of the multi-agent workflow through a graphical interface and programming interface, and embed open-source tools and self-developed tools in the workflow to achieve modular processing of tasks. The user selects multiple agents for combination in the workflow editing interface of the workbench. The workflow configuration module automatically generates task assignment logic and allows the user to customize and optimize it. At the same time, it supports embedding external tools in the workflow. The external tools include GPT, Excel analysis tools, etc.;

[0025] The publishing and integration module is used to publish the configured agent and workflow to the cloud for other users to use. The publishing and integration module optimizes the agent and workflow through the ChainGen framework, and at the same time supports embedding into other agents and workflows to achieve seamless cross-platform integration. Through the publishing and integration module, the configured agent and workflow can be published to the cloud, and the workbench provides an API interface to support the docking of the agent with other systems;

[0026] The WeCom access module is used to support the access of the agent to WeCom, realize the collection and analysis of WeCom chat records, and at the same time automatically incorporate the chat records into the agent knowledge base to improve the knowledge management ability. Through the WeCom access module, the user accesses the agent through WeCom or other applications to realize an intelligent work process.

[0027] The workbench supports the combination of multiple agents and can embed open-source tools and self-developed tools.

[0028] The workbench collects chat records through the enterprise WeChat access module and saves them to the intelligent agent knowledge base.

[0029] When applied to the intelligent transformation of traditional software enterprises, it can help traditional software enterprises quickly achieve Agentization and improve operational efficiency;

[0030] When applied to enterprise-level intelligent agent innovation, it pioneered an Agentization solution for the enterprise-level market, filling the industry gap;

[0031] When applied to Agent, it created the first Agentized company in the RMN market, achieving an industry benchmark;

[0032] When applied to intelligent business analysis (ChatBI), it provides Agentization support for business analysis and realizes data-driven decision-making;

[0033] When applied to AI content creation (AIGC), it supports the Agentization of content generation tasks and realizes efficient creative generation and optimization.

[0034] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. An agent building workbench based on a multi-agent framework, characterized in that: The workbench includes an agent building module, a workflow configuration module, a publishing and integration module, and an enterprise WeChat access module; The agent building module is used to quickly create an agent according to business needs, and configure the parameters, functional modules and knowledge base of the agent; The workflow configuration module builds a task allocation mechanism based on the open source multi-agents architecture. The workflow configuration module is used to support the configuration of multi-agent workflows through a graphical interface and a programming interface, and embed open source tools and self-developed tools in the workflow to achieve modular processing of tasks; The publishing and integration module is used to publish the configured agents and workflows to the cloud for use by other users. The publishing and integration module optimizes agents and workflows through the ChainGen framework and supports embedding into other agents and workflows to achieve seamless integration across platforms. The enterprise WeChat access module is used to support the access of intelligent entities to enterprise WeChat, realize the collection and analysis of enterprise WeChat chat records, and automatically incorporate the chat records into the intelligent entity knowledge base to improve knowledge management capabilities.

2. An agent building workbench based on a multi-agent framework as claimed in claim 1, characterized in that: The workbench supports multi-agent combination and can embed open source tools and self-developed tools.

3. An agent building workbench based on a multi-agent framework as claimed in claim 1, characterized in that: The workbench collects chat records through the enterprise WeChat access module and saves them to the intelligent body knowledge base.

Images