Chi Wang

AutoGen is an open-source multi-agent conversation framework developed by Microsoft Research that enables developers to build agentic systems where multiple LLM-powered agents, tools, and human participants interact through a shared conversation protocol to complete complex tasks. The framework supports customizable agent roles, automated code execution, tool use, and human-in-the-loop workflows.\n\nThe project is maintained by Microsoft Research and has accumulated one of the largest GitHub star counts of any AI agent framework, reflecting broad adoption among researchers and engineers exploring multi-agent system architectures. AutoGen is widely used in academic research on agent coordination and in production systems at organizations building AI-powered automation for software engineering, data analysis, and business process tasks.\n\nMulti-agent AI frameworks represent a foundational research and engineering area as the field moves toward systems where many specialized AI agents collaborate on tasks requiring diverse capabilities. AutoGen backing by Microsoft Research gives it institutional credibility and ensures continued development investment, and its influence on production multi-agent system design has been substantial, with many commercial multi-agent products drawing on architectural patterns established in the AutoGen framework.