What we covered
ADK builds the agents; A2A gives them a common language. The session treats them as one story, because an agent that can't talk to anything outside its own framework is a feature, not a system.
- What AI agents are, and the challenges that make building them harder than they look.
- ADK as an agent building toolkit: its key features, and a short code example.
- A walkthrough of an ADK agent, from definition to running.
- A2A next to MCP — the comparison everyone asks for, and where each protocol belongs.
- How A2A works: the interaction flow, its core concepts, and how responses and updates come back.
- ADK and A2A together, and what the combination makes possible that neither does alone.
- Use cases across industries, and what this means concretely for developers.
- Next steps and resources.
Format
A technical overview talk with code examples and short walkthroughs of both an ADK agent and an A2A interaction.
Who it's for
Developers and architects evaluating agent frameworks, particularly anyone who expects to run agents built by more than one team.