AI2Apps: A Visual IDE for Building LLM-based AI Agent Applications

Xin Pang,Zhucong Li,Jiaxiang Chen,Yuan Cheng,Yinghui Xu,Yuan Qi

2024-04-07

Abstract:We introduce AI2Apps, a Visual Integrated Development Environment (Visual IDE) with full-cycle capabilities that accelerates developers to build deployable LLM-based AI agent Applications. This Visual IDE prioritizes both the Integrity of its development tools and the Visuality of its components, ensuring a smooth and efficient building experience.On one hand, AI2Apps integrates a comprehensive development toolkit ranging from a prototyping canvas and AI-assisted code editor to agent debugger, management system, and deployment tools all within a web-based graphical user interface. On the other hand, AI2Apps visualizes reusable front-end and back-end code as intuitive drag-and-drop components. Furthermore, a plugin system named AI2Apps Extension (AAE) is designed for Extensibility, showcasing how a new plugin with 20 components enables web agent to mimic human-like browsing behavior. Our case study demonstrates substantial efficiency improvements, with AI2Apps reducing token consumption and API calls when debugging a specific sophisticated multimodal agent by approximately 90% and 80%, respectively. The AI2Apps, including an online demo, open-source code, and a screencast video, is now publicly accessible.

Artificial Intelligence,Software Engineering

What problem does this paper attempt to address?

This paper introduces a visual integrated development environment (Visual IDE) called AI2Apps, aimed at accelerating developers in building AI agent applications based on large language models (LLMs). Existing development tools and platforms currently have limitations in terms of programming flexibility, debugging efficiency, and visualization. AI2Apps addresses these issues by providing a complete development toolkit, including a prototype design canvas, AI-assisted code editor, agent debugger, management system and deployment tools, as well as an extensible plugin system (AI2Apps Extension, AAE).

The main features of AI2Apps include:
1. Completeness: It provides a web-based graphical user interface that integrates various development tools such as prototype design, code editing, debugging, and deployment, thus improving programming efficiency.
2. Visualization: It converts frontend and backend code into intuitive drag-and-drop components, simplifying the development process and providing over 50 GUI widgets to enhance user interaction.
3. Extensibility: Developers can create new functional components, such as new plugins that simulate human browsing behavior, using the AAE plugin system and open technologies.

The case studies mentioned in the paper demonstrate that AI2Apps reduces token consumption by approximately 90% and API calls by 80% when debugging complex multimodal agent applications, significantly improving efficiency.

In summary, the goal of AI2Apps is to address the efficiency and visualization challenges in LLM-based AI agent application development, providing a comprehensive and intuitive development environment to facilitate the rapid construction and deployment of such applications.