Description
Most AI agents are a single LLM wrapped in a system prompt. They work in demos, then collapse in production. The fix is specialized agents that coordinate — each armed with MCP tools to act on the real world.
In this live-coding session, you’ll see a multi-agent system built with Google’s ADK and MCP servers in Python, covering three orchestration patterns: sequential, parallel, and iterative. ADK gives you control flow. MCP gives each agent its capabilities. You’ll leave with working code and patterns ready for production.
Speaker
Muhammad Ahsan Ayaz
Software Architect, Scania