aiml
SearchMy learning
Menu

Google ADK 2.0 learning path

Google ADK 2.0 courses

Start with Building Agents, then choose the courses your application needs. The four courses cover agent tools and workflows, deployment, live conversations, and service integrations.

Courses, modules and lessons

  1. Course 1

    Building Agents with Google ADK 2.0

    Build agents, tools and multi-step workflows with Google ADK 2.0, then evaluate them.

    17 lessons · 3 modules · about 6 hours of reading

    Prerequisites: Python and HTTP APIs.

    Module 1: Agents and tools6 lessons

    Full module overview

    1. Build your first ADK agent with Gemini
    2. Run an agent and read its event stream
    3. Give an agent typed function tools
    4. Connect an agent to MCP servers with McpToolset
    5. Persist session state and long-term memory
    6. Guard model and tool calls with callbacks and plugins

    Module 2: Workflows and Evals6 lessons

    Full module overview

    1. Build a graph workflow from nodes and edges
    2. Route a workflow with conditional edges
    3. Pause a workflow for human approval
    4. Build dynamic workflows in Python
    5. Delegate to sub-agents through a coordinator
    6. Evaluate an agent's tool trajectory and response

    Module 3: Workflow patterns5 lessons

    Full module overview

    1. Choose an ADK workflow pattern
    2. Generate and review
    3. Iterative refinement
    4. Parallel fan-out and gather
    5. Hierarchical task decomposition

    Course overviewStart course

  2. Course 2

    Running and Deploying Agents with Google ADK 2.0

    Serve, observe, evaluate and deploy ADK agents with runtime controls, context caching, traces, simulated users, Cloud Run and Agent Runtime.

    20 lessons · 5 modules · about 8 hours of reading

    Prerequisites: Python, HTTP and JSON, and the Building Agents with Google ADK 2.0 course.

    Module 1: Run agents and manage sessions6 lessons

    Full module overview

    1. Serve an agent over HTTP and drive it from the CLI
    2. Configure an ADK Agent with YAML
    3. Tune a run with RunConfig and read the event loop
    4. Resume an interrupted agent run
    5. Rewind a session and migrate its database
    6. Run an agent from event triggers

    Module 2: Manage context and artifacts2 lessons

    Full module overview

    1. Keep long sessions in budget with compaction and context caching
    2. Save and load files with Artifacts

    Module 3: Connect APIs and tools3 lessons

    Full module overview

    1. Call a REST API from an OpenAPI spec, with authentication
    2. Make tools parallel-ready
    3. Expose the docs assistant as an MCP server

    Module 4: Evaluate, trace and apply guardrails5 lessons

    Full module overview

    1. Trace and measure agent runs with OpenTelemetry
    2. Test an agent against simulated users
    3. Inject tool failures with environment simulation
    4. Write a custom eval metric and optimize the instruction
    5. Apply safety and security guardrails to agent tools

    Module 5: Deploy agents on Google Cloud4 lessons

    Full module overview

    1. Use Agent Platform models, sessions and Memory Bank
    2. Scaffold and deploy an agent with the Agents CLI
    3. Deploy an ADK agent to Agent Runtime
    4. Deploy an ADK agent to Cloud Run

    Course overviewStart course

  3. Course 3

    Building Live and Voice Agents with Google ADK 2.0

    Stream audio and screen frames with ADK live agents, handle tools and sessions, test voice callbacks, and build a custom WebSocket bridge.

    7 lessons · 3 modules · about 3 hours of reading

    Prerequisites: Python async code and the Building Agents with Google ADK 2.0 course or equivalent ADK experience. Live runs need a Gemini Live API key; recorded replays do not.

    Module 1: Stream media and manage live sessions3 lessons

    Full module overview

    1. Stream a voice conversation with run_live
    2. Read live events and manage live sessions
    3. Send audio, images and video, and configure voice

    Module 2: Call tools and run live workflows2 lessons

    Full module overview

    1. Call tools during a live conversation
    2. Run live agents inside a workflow

    Module 3: Test and serve live agents2 lessons

    Full module overview

    1. Guard and evaluate a voice agent
    2. Build a custom WebSocket server for a live agent

    Course overviewStart course

  4. Course 4

    Connecting Agents to Tools and Services with Google ADK 2.0

    Connect agents to BigQuery, databases through MCP Toolbox, Agent Search, Google Search and code execution, then add Skills and A2A.

    6 lessons · 2 modules · about 2 hours of reading

    Prerequisites: Python, HTTP APIs, and ADK agents with tools (the Building Agents with Google ADK 2.0 course covers them).

    Module 1: Add Skills and built-in Gemini tools2 lessons

    Full module overview

    1. Give an agent skills from a folder with SkillToolset
    2. Use Gemini built-in tools: Google Search grounding and code execution

    Module 2: Connect databases, search and other agents4 lessons

    Full module overview

    1. Query PostgreSQL through MCP Toolbox for Databases
    2. Query BigQuery with BigQueryToolset
    3. Search private data stores with Agent Search
    4. Expose the docs assistant over A2A and consume it

    Course overviewStart course