Google ADK 2.0

The Google ADK 2.0 course is designed to help learners build practical skills in developing, orchestrating, evaluating, and deploying AI agents using Google's Agent Development Kit (ADK). Through conceptual lessons, guided demonstrations, and hands-on labs, you'll learn how to build your first agent, configure instructions and tools, and understand the core components that make up an ADK-based agent.

The course explores ADK's workflow runtime, including workflow graphs, conditional routing, parallel and looping agents, dynamic fanout, retries, and human-in-the-loop approval. You'll also learn how to manage session state, use context compaction, and work with artifacts and memory to build more capable and context-aware agents.

As you progress, you'll explore advanced tool integrations using MCP and OpenAPI, tool execution confirmation, agent evaluation, and API integration. The course also covers deploying agents to Google Cloud Run, helping you understand how to take your agent applications from development to deployment.

Course Modules & Learning Outcomes

Fundamentals of Google ADK

Build a strong foundation in Google's Agent Development Kit and understand the capabilities introduced in version 2.0. Learn how to set up your development environment, build your first agent, and understand its core anatomy. You'll also learn how to configure agent instructions, integrate tools, and provide your agent with capabilities through a hands-on lab.

ADK's Workflow Runtime

Learn how to design and orchestrate multi-agent workflows using ADK's workflow runtime. Explore workflow graphs, conditional routing, parallel execution, looping agents, and the differences between static and dynamic workflows. You'll also learn about dynamic fanout, retries, and human-in-the-loop approval, applying these concepts through hands-on workflow and approval labs.

Memory and State Management

Understand how session state and memory support context-aware agent interactions. Learn about context compaction, artifacts, and memory management, and apply these concepts to build agents that can maintain and manage information across interactions.

Advanced Tools

Explore advanced capabilities for extending and operating ADK agents. Learn how to integrate MCP and OpenAPI tools, implement tool execution confirmation, and evaluate agent performance. You'll also learn how to expose agents through APIs and deploy them to Google Cloud Run.

Bringing It All Together

Reinforce the concepts covered throughout the course through a comprehensive course quiz. Review the key principles of agent development, workflow orchestration, state management, advanced tools, evaluation, and deployment.

Course Features

• Hands-on labs for building agents, integrating tools, creating workflow graphs, and implementing human-in-the-loop interactions.
• Practical experience with ADK 2.0's workflow runtime, including conditional routing, parallel execution, looping, dynamic fanout, and retries.
• Comprehensive coverage of session state, context compaction, artifacts, and memory management.
• Advanced tool integrations using MCP and OpenAPI, along with tool confirmation and agent evaluation.
• Practical guidance on exposing agents through APIs and deploying agent applications to Google Cloud Run.

Who Should Enroll?

• AI engineers and developers building agentic applications.
• Software developers interested in Google's Agent Development Kit.
• Cloud engineers developing and deploying AI-powered solutions on Google Cloud.
• DevOps and platform engineers supporting AI agent infrastructure and deployment.
• Developers working with workflow orchestration, tool integrations, and multi-agent systems.
• Anyone looking to build practical skills in developing and deploying AI agents using Google ADK 2.0.

Build practical expertise in Google ADK 2.0 and learn how to develop, orchestrate, extend, evaluate, and deploy AI agents using workflow automation, advanced tools, memory management, and Google Cloud Run.