Deployment & Scaling · Advanced
Agents on Google Cloud: ADK, Agent Runtime and Agent2Agent
The Agent Development Kit for writing agents in code, Agent Runtime (formerly Agent Engine) for hosting them with sessions and memory, and the interoperability protocols that let agents talk to tools and to each other.
- 1.6 hours lesson
- advanced level
- Google Cloud AI Track
- Narrated presentation
What you’ll learn
- ADK: open-source, model-agnostic, Python and Java
- LlmAgent, Sequential/Parallel/Loop workflow agents, and sub-agent delegation
- The ADK dev UI event stream as the primary debugging tool
- Agent Runtime with Sessions, Memory Bank and sandboxed Code Execution
- Cloud Run as the control-first alternative to managed Agent Runtime
- MCP for tool interoperability, A2A for agent-to-agent collaboration
- Turn caps, parallel tools, tracing and trajectory evaluation as the real failure controls
Inside this lesson
A narrated, slide-by-slide presentation with 1 prediction exercise, 1 knowledge check, 4 videos. You answer as you go, and the platform checks your work.