Build AI Agents using Google Agent Development Kit (ADK)

Name: Build AI Agents using Google Agent Development Kit (ADK)
Rating: 4.3 (113 reviews)

Develop Your First AI Agent in 2 Hours. Get Started with Google ADK with Hands on Experience

Created byK8s Point - Training House for GCP, AI & Kubernetes

Last updated 6/2026

English

English [Auto],

What you'll learn

Start your AI Agent journey with a Free Course
Agentic AI solutions from Google
2 Hours and You will be ready with your first AI Agent
Easy to learn material and all codes and commands included
Deploy your agent in Vertex AI Agent Engine with simple command

Course content

4 sections • 25 lectures • 1h 58m total length

Course Introduction2:22
Introduction to AI Agents8:04
Explore how AI agents with Google ADK outperform traditional LMs by using a brain-like decision process, coordinating weather checks, flight and hotel searches, and bookings via RAG operations and tools.
Introduction to Google ADK4:52
Explore how Google ADK enables AI agents within the Google cloud ecosystem, offering multi-modal support, tool integrations, and deployment via Vertex AI, Cloud Run, or GKE.
Documentation and Example Links2:02
Explore the SDK documentation and GitHub examples to learn agent development, with API references for Python, Go, and Java, and steps to clone and run your agent from your laptop.
Important Announcement on Gemini 2.5 Flash Model0:30

Download the Programs Used0:01
Install ADK Library & Understand the Directory Structure7:47
Install the adk library and verify you are on sdk 1.16 or newer, then use sdk create to build a folder structure with env, __init__.py, and agent.py.
Create Gemini API Key for the Agent6:16
Learn to create a Gemini API key using Google AI Studio, link it to a GCP or non-GCP project, and use it with the agent.
Setup Vertex AI as Backend for the Agent4:23
Create a Vertex AI agent and set Vertex AI as the backend, using the same model as the flash model, then enable the Vertex AI API for the project.
Understand the Agent Definition5:16
Define the first agent with the LM agent class, name route agent, and configure model, instruction, and description, then use history agent to produce bullet-point summaries and a pipeline summary.
Run Your Agent with Command - adk web5:20
Test your agent locally with sdk web, the run command, and sdk api server before deployment. Use the edq web interface to interact and debug the history agent.
Run Your Agent with Command - adk run2:28
Test your agent with SDK run, deploying a command line interface to interact with the agent using World War Two as an example, and compare CLI to the web interface.
Run Your Agent with Command - adk api_server8:10
Run your agent via the API server, exposing a restful interface accessible over the network, with curl-based sessions to list apps and interact in real time.
Revisit the ADK Web Interface2:52
Revisit the ADK web interface to access the agent home page, view session IDs, and manage sessions. Speak via microphone, attach files, or use the camera to interact and debug.
Run Your Agent from Your Desktop/Laptop - API Key as Backend5:32
Set up a local desktop agent project with PyCharm, install Google SDK, and configure a history agent using a Gemini API key as backend, running via SDK web on localhost.
Run Your Agent from Your Desktop/Laptop - Vertex AI as Backend3:05
Run your agent from a desktop using Vertex AI as the backend, configure project and location, and authenticate with gcloud application default login or an API key.
Using Service Account to Authenticate an Agent (LLM)10:38
Learn to authenticate an agent with a Google service account for Vertex AI, assign the Vertex AI user role, and apply workload identity federation and secret manager best practices.
Non Google Model with ADK ( Ex - Claude or OpenAI)5:16
Learn to use non-google models with the ADK, authenticating with API keys and selecting third-party LLMs like Anthropic or OpenAI. Configure endpoints from Vertex AI to self-hosted to run models.
Agent Designer - UI Option7:47
Discover the preview of agent designer in Google Agent Development Kit (ADK), a UI-driven tool to create a root agent with sub agents, add tools, generate code, and test interactively.

Download the Artifacts0:02
Agent Deployment Options4:20
Explore agent deployment options in Google Cloud, including Vertex AI agent engine, Cloud Run, and GKE cluster, and learn when to use each for production, with memory bank support.
Deploy ADK Agent with in Vertex AI Agent Engine7:24
Deploy your ADK agent to Vertex AI Agent Engine via edk deploy using a staging bucket, and name the directory with underscores (not hyphens); monitor logs for issues.
Test the Agent with curl Command8:05
Test a deployed agent with curl by creating a session and sending rest queries, using the agent engine console, query URLs, and stream queries, while noting IAM access in GCP.
Test the Agent from Python Program5:43
Learn how to access an agent from a Python program using REST URLs, including authentication, session creation, and optional streaming responses, demonstrated in Colab Enterprise.

Requirements

Very Basic Python
Eager to become AI Agents Champion

Description

Artificial Intelligence agents are transforming how modern applications are built. In this beginner-friendly course, you will learn how to create your very first AI agent using the Google Agent Development Kit (ADK) and deploy it using Agent Engine on Google Cloud.

This course is designed for developers, cloud engineers, and AI enthusiasts who want a practical introduction to building intelligent agents. Instead of focusing only on theory, we will walk through a hands-on example where you will build an AI agent step by step and see how it can be deployed to run in the cloud.

By the end of this course, you will understand how AI agents work, how to build one using Google ADK, and how to deploy it using Google Agent Engine so it can run as a real application. This course is perfect if you want to quickly get started with modern AI agent development on Google Cloud.

If you are curious about the future of AI-driven applications and want to learn how to build and deploy your first AI agent, this course will give you the perfect starting point.

Build your AI Agents with Google Agent Development Kit (ADK).

A mini course on Google ADK with step by step process on how to develop and deploy your Agent in Google Cloud.

Environment Setup ( Google Cloud / Local Desktop - laptop)
Authenticate using Gemini Key / Vertex AI / 3rd Party ( Open AI / Anthropic)
Learn to test your agent locally with adk web, adk run and adk api_server options
Use built-in tools, custom tools, 3rd party tools
Use CrewAI and Langchain tools with ADK AI Agents

Who this course is for:

Anyone who wants to become AI Agent Developer
Everyone who is in the world of AI

Build AI Agents using Google Agent Development Kit (ADK)

What you'll learn

Explore related topics

Course content

Introduction5 lectures • 18min

Develop Your First Agent14 lectures • 1hr 15min

Deployment of Your First Agent5 lectures • 26min

What Next!1 lecture • 1min

Requirements

Description

Who this course is for: