AutoGen

AutoGen is an open-source framework from Microsoft for building multi-agent AI systems that coordinate large language models (LLMs), tools, and humans to solve complex tasks. It provides a structured way to define multiple specialized agents, configure their roles and capabilities, and orchestrate their interactions through automated conversations. Developers can create agents that use different LLMs, call external tools and APIs, execute code, retrieve data, or request human input when needed. AutoGen supports both fully automated workflows and human-in-the-loop scenarios, making it suitable for production systems that require reliability and oversight.

Key features include customizable agent classes, conversation management, function and tool calling, and built-in support for code execution and validation. The framework simplifies common patterns such as task decomposition, iterative refinement, and collaborative problem-solving between agents. Typical use cases include automated data analysis, code generation and debugging, multi-step research assistants, workflow automation, and domain-specific copilots that combine several specialized agents. AutoGen also offers examples, templates, and integration patterns to help teams quickly prototype and deploy agent-based applications while maintaining control over configuration, security, and resource usage.