Multi-Agent AI Systems

Agent-Framework - Unified Platform for A2A Agents

Hands-on lab notes from LAB513 Series: Agent-framework for multi-agent systems.

Coming Soon

This article will be published during/after Microsoft Ignite 2025 (Nov 18-21). Full hands-on walkthrough of building A2A agents with the unified framework coming soon.


Lab Overview

What We're Building:

  • A2A (Agent-to-Agent) compatible agents using agent-framework
  • SWE Agents with GitHub Copilot coding agent and Codex
  • MCP tools for secure enterprise workflows
  • Multi-agent orchestration similar to magnetic-one

Technologies:

  • Agent-framework (unified Semantic Kernel + AutoGen platform)
  • Azure OpenAI models
  • MCP (Model Context Protocol)
  • Pre-release capabilities (embargoed content)

Key Learning Goals

  1. Agent-framework fundamentals - How does the unified platform work?
  2. A2A protocol - What does Agent-to-Agent communication look like?
  3. SWE Agents - How do coding agents accelerate development?
  4. MCP security - Enterprise-safe tool integration patterns
  5. Deployment - Going from lab to production

Stay Tuned

Full lab walkthrough, code samples, and architectural insights coming soon.

Sessions: LAB513 + LAB513-R1/R2/R3 | Various times Nov 19-21 | Moscone West

Previous
Building with Azure AI Foundry
Built: Dec 16, 2025, 05:16 AM PST
b8041d3