Skip to main content

Google ADK Masterclass Roadmap

Quick Start

Ready to start learning?

Welcome to the complete curriculum for the Google Agent Development Kit (ADK) Masterclass. This course is designed to take you from a local development environment to deploying enterprise-grade, multi-agent systems on Google Cloud.

Bestseller
Master Google ADK - Agent Development Kit

Master Google ADK - Agent Development Kit

Build Multi Agent Systems | Mac, Windows, Ubuntu | Deploy to Google Cloud | Free Gemini Key

★ 4.4👥 470+
View on Udemy

The roadmap below outlines the 9 core pillars of the course, comprising 87 detailed lectures.

📚 Detailed Curriculum​

Section 1: Foundations of ADK​

Start Here. We dissect the Google ADK framework, understanding why it differs from other frameworks like LangChain or CrewAI, and explore the interoperability features that make it unique.

  • Lecture 1: About the Course & Instructor
  • Lecture 2: Introduction & What are Agent Frameworks?
  • Lecture 3: Why ADK? How is ADK Different?
  • Lecture 4: ADK UI Playground
  • Lecture 5: ADK Interoperability
  • Lecture 6: Summary, Important Points to Note!

Section 2: Environment Setup & First Agent​

The "Hello World" of Agents. We set up a professional AI engineering environment using UV (the modern Python package manager), VS Code, and Git across all major operating systems.

  • Lecture 7: Introduction and Quick Preview of What We'll Build
  • Lecture 8: How to Build This Agent
  • Lecture 9: [macOS] Install Python, UV, VS Code and Git
  • Lecture 10: [Windows] Install Python, UV, VS Code and Git
  • Lecture 11: [Ubuntu] Install Python, UV, VS Code and Git
  • Lecture 12: How to Access the Course Code
  • Lecture 13: [macOS] Setup Code and Run
  • Lecture 14: [Windows] Setup Code and Run
  • Lecture 15: [Ubuntu] Setup Code and Run
  • Lecture 16: [macOS, Windows, Ubuntu] Setup Code and Run
  • Lecture 17: [macOS, Windows, Ubuntu] Code walkthrough

Section 3: Sequential Workflow Agents​

The Multi-Agent Pipeline. In this section, we live-code a complex Website Builder. You will learn how to chain agents together where the output of one becomes the context for the next.

  • Architecture: Requirements Agent $\rightarrow$ Designer Agent $\rightarrow$ Code Writer Agent.

  • Lecture 18: Introduction and Overall Steps

  • Lecture 19: STEP 1 - Define Agent Architecture

  • Lecture 20: STEP 2 - Setup Folders

  • Lecture 21: STEP 3.1 - Coding the Requirements Writer Agent | Types of ADK Agents

  • Lecture 22: STEP 3.2 - Coding the Designer Agent

  • Lecture 23: STEP 3.3 - Coding the Code Writer Agent

  • Lecture 24: STEP 3.4 - Coding the Root Agent (Sequential Workflow Agent)

  • Lecture 25: STEP 3.5 - Environment Setup

  • Lecture 26: STEP 4 - Testing the Agent

  • Lecture 27: Let's look at the Agent Output!

  • Lecture 28: Summary

Bestseller
Master Google ADK - Agent Development Kit

Master Google ADK - Agent Development Kit

Build Multi Agent Systems | Mac, Windows, Ubuntu | Deploy to Google Cloud | Free Gemini Key

★ 4.4👥 470+
View on Udemy

Section 4: ADK Execution Methods​

Runtime Versatility. Master the four distinct ways to invoke and interact with your ADK agents, giving you flexibility for any integration scenario.

  • Lecture 49: Quick Preview of the four methods
  • Lecture 50: Method 1: ADK WEB
  • Lecture 51: Method 2: ADK API SERVER
  • Lecture 52: Method 3 (Part 1): ADK Runner Program - Code Overview
  • Lecture 53: Method 3 (Part 2): ADK Runner Program - Agent Execution
  • Lecture 54: Method 4: ADK RUN Command Line Interface
  • Lecture 55: Summary

Section 5: Model Context Protocol (MCP)​

Universal Integration. This is the cutting edge of AI. We explore the Model Context Protocol (MCP) to build agents that can connect to any data source or tool in a standardized way. We build a custom MCP Client and Server.

  • Lecture 56: Why Build MCP Clients with Google ADK?
  • Lecture 57: STEP 1 - Building MCP Servers for our ADK based MCP Client
  • Lecture 58: STEP 2 Building the ADK MCP Client
  • Lecture 59: Introduction, Preview of What We'll Build
  • Lecture 60: Code overview
  • Lecture 61: GOOGLE_API_KEY setup
  • Lecture 62: MCP Client User Interface - cmd.py
  • Lecture 63: MCP Client Implementation - client.py
  • Lecture 64: MCP Agent Implementation - agent.py
  • Lecture 65: Code - MCP config json file
  • Lecture 66: Code for Utilities & STDIO Server + Environment Setup
Bestseller
Master Google ADK - Agent Development Kit

Master Google ADK - Agent Development Kit

Build Multi Agent Systems | Mac, Windows, Ubuntu | Deploy to Google Cloud | Free Gemini Key

★ 4.4👥 470+
View on Udemy

Section 6: Parallel Agent Architectures​

High-Performance AI. Learn to break monolithic tasks into sub-tasks that run simultaneously. We build a Research Suite that generates questions, researches them in parallel, and synthesizes the results.

  • Lecture 29: Parallel Agent Architecture and Code Structure
  • Lecture 30: Root Sequential Agent
  • Lecture 31: Questions Generator LLM Agent
  • Lecture 32: Questions Researcher Parallel Agent
  • Lecture 33: Query Generator LLM Agent
  • Lecture 34: Review of Agents from previous lectures
  • Lecture 35: Agent Setup and Demo

Read the Documentation →

Section 7: Deploy to Google Cloud (Coming Soon)​

Docs In Progress

The written documentation for Google Cloud Deployment is currently being authored.

Production Engineering. Move your agents off localhost. We cover the complete DevOps pipeline: Dockerizing your agent, setting up FastAPI, managing secrets, and deploying to Google Cloud Run via the gcloud CLI.

  • Lecture 36: Introduction and Agent Architecture
  • Lecture 37: Getting the code
  • Lecture 38: Clone or update your code from the adk_samples Repository
  • Lecture 39: Uvicorn server and fast api app
  • Lecture 40: Dockerfile
  • Lecture 41: CLOUD SETUP - Create a new Gmail Account if you prefer
  • Lecture 42: CLOUD SETUP - Create a Google Cloud Project
  • Lecture 43: CLOUD SETUP - Install gcloud CLI
  • Lecture 44: CLOUD SETUP - Enabling API Keys if needed
  • Lecture 45: Environment Variables File
  • Lecture 46: Deploy to Google Cloud Run
  • Lecture 47: Demo of Agent on Google Cloud Run
  • Lecture 48: Cleanup (important to avoid charges) and conclusion

Section 8: No-Code Agents (YAML) (Coming Soon)​

Docs In Progress

The written documentation for No-Code Agents is currently being authored.

Declarative AI Engineering. Learn to architect complex multi-agent systems without writing Python code for the orchestration layer. We use YAML configurations to define agent behaviors and relationships.

  • Lecture 67: Introduction to No-code Agents and YAML Config
  • Lecture 68: Setup your environment
  • Lecture 69: ADK Create & Get API Key
  • Lecture 70: Agent 1 - root agent
  • Lecture 71: Multi-Agent Orchestration
  • Lecture 72: Sub Agents - physics tutor
  • Lecture 73: Sub Agent - python tutor
  • Lecture 74: Putting it all together
  • Lecture 75: Running The Agent

Section 9: ADK Visual Agent Builder (Coming Soon)​

Docs In Progress

The written documentation for Visual Agent Builder is currently being authored.

Rapid Prototyping. We conclude by exploring the visual side of ADK. Use a drag-and-drop canvas to design agent flows and export them to code.

  • Lecture 76: Introduction
  • Lecture 77: UI Overview
  • Lecture 78: Caveats
  • Lecture 79: What we are going to build
  • Lecture 80: Setup
  • Lecture 81: Get API Key
  • Lecture 82: Start Agent Builder
  • Lecture 83: Create App
  • Lecture 84: Create Agent
  • Lecture 85: Test your Agent
  • Lecture 86: Agent Builder Assistant
  • Lecture 87: Summary
Bestseller
Master Google ADK - Agent Development Kit

Master Google ADK - Agent Development Kit

Build Multi Agent Systems | Mac, Windows, Ubuntu | Deploy to Google Cloud | Free Gemini Key

★ 4.4👥 470+
View on Udemy
Ready to Start?

The roadmap above represents a journey from Beginner to AI Architect. By the end of Lecture 87, you will have a portfolio of deployed, functional, and scalable agents.