Learn Microsoft AI explains the Microsoft Agent Framework, highlighting its evolution from Semantic Kernel and AutoGen, and its advantages for building robust AI agent systems with .NET and Python.

What Is Microsoft Agent Framework & Why Another Agent Framework?

In this video, Learn Microsoft AI explains the purpose, structure, and benefits of the Microsoft Agent Framework—the latest open-source offering from Microsoft for developing LLM-powered agents and multi-agent workflows.

Overview

  • Microsoft Agent Framework is an open-source AI development kit designed to help developers build agents and multi-agent orchestration workflows using .NET and Python.
  • It unifies and extends the capabilities of its predecessors: Semantic Kernel (enterprise feature focus) and AutoGen (agent abstractions and multi-agent experimentation).

Key Topics Covered

  • Framework Structure: Breaks down how Microsoft Agent Framework is architected for modularity, including LLM integrations, tool use, and workflow orchestration.
  • Enterprise and Agent Features: Inherits state management, type safety, telemetry, and broad model support from Semantic Kernel, while adding explicit graph-based workflows from AutoGen.
  • Multi-Agent Orchestration: Supports graph workflows, checkpointing, and human-in-the-loop scenarios, suitable for complex enterprise applications.
  • Transition from Previous Frameworks: Explains why Microsoft combined Semantic Kernel and AutoGen, creating a single, production-ready foundation for modern AI agent systems.
  • Real-world Use Cases: Demonstrates how developers can build robust, long-running agent workflows with enhanced observability and control.

Why Choose Agent Framework?

  • Seamlessly integrates enterprise-grade features with flexible agent abstractions.
  • Facilitates building scalable, maintainable AI agent systems for both experimentation and production scenarios.
  • Supports cross-language development with .NET and Python.

Learn More

For further resources and discussion, see the video description for social and community links.