Session abstract
What you’ll learn
Everyone's building AI agents. But most of them can't talk to each other. They're islands. In this session, we'll walk through a real multi-agent system built with Google's Agent Development Kit and the Agent2Agent (A2A) protocol: a wardrobe recommendation app that pulls live weather data, reasons about what to wear, and coordinates across distributed services. You'll see how A2A enables agents to discover each other's capabilities, communicate across deployments, and scale independently, all without tight coupling. We'll also cover how real-time data pipelines (via Conduit) feed your agents, and how to deploy the whole system to Cloud Run. Whether you're building internal tooling or consumer-facing AI features, this talk gives you a battle-tested blueprint for multi-agent architecture that goes beyond the hype.
