AI Agents with Google's ADK Bootcamp: Build with Gemini, Protocols & AgentOps (SED)
- • Nov 16–18, 2026 · Live Remote Online
Contact IT Dojo for current pricing, available dates, and a custom quote tailored to your team or organization.
Course Duration
3 Half-Day Sessions (13.5 Hours)
Audience
Employees of federal, state and local governments; and businesses working with the government.
Prerequisites
Basic Python proficiency, familiarity with cloud concepts (APIs, services), an understanding of core LLM concepts such as prompting and RAG, and comfort at the command line. Students need a computer with Python and a code editor (VS Code recommended) and a Google Cloud account with billing enabled (free credits are available).
Course Description
This hands-on bootcamp takes you from zero to launch-ready AI agents with Google's Agent Development Kit (ADK). You will set up ADK, run your first agents, connect Gemini and other models through LiteLLM, define custom Python tools with proper schemas, and add conversational memory. From there you will apply agentic design patterns such as RAG and sub-agent delegation, make agents interoperable with the Model Context Protocol (MCP) and the Agent-to-Agent (A2A) protocol, and finish with evaluation, AgentOps, deployment to Vertex AI Agent Engine, and agent security practices including prompt injection defenses and guardrails. Every session blends presentation, live demonstration, hands-on exercises, and group discussion.
Learning Objectives
- Explain the core concepts and workflow of building agents with ADK
- Run and test agents with the ADK command-line tools
- Connect agents to Gemini and to open models via LiteLLM
- Define and register custom tools using Python functions
- Implement conversational memory and choose persistence strategies
- Apply multi-agent systems and agentic design patterns
- Use MCP and A2A to make agents interoperable
- Evaluate, deploy, and secure agents for production
Course Outline
1. Introduction to Agents and ADK
- What AI agents are and why they matter now
- Core ADK concepts: agent, tool, memory, workflow
- The agent framework landscape: ADK, LangChain, CrewAI, LangGraph
- Hands-on: create a sample agent with adk init and test it in adk web
2. Models and Tools
- Connecting ADK to Gemini on Vertex AI and choosing a model
- Connecting to open models with LiteLLM
- Defining Python tools, schemas, and parameter types
- Hands-on: register a custom tool in your agent
3. Multimodal Agents and Memory
- Multimodal input with Gemini
- Memory types, persistence, artifacts, and state
- Hands-on: add memory and observe how context changes responses
4. Agentic Design Patterns
- Retrieval-augmented generation with ADK
- Sequential, delegate, and loop sub-agent patterns
- Single-agent vs. multi-agent architectures
- Hands-on: implement a design pattern for your use case
5. Putting It All Together
- The ADK agent lifecycle from development to deployment
- Hands-on: build a multi-tool daily briefing agent
6. MCP and A2A
- Model Context Protocol for tools and services
- Agent-to-Agent communication standards
- Hands-on: swap in an MCP server tool and make your agent A2A-compatible
7. Evaluation, Deployment, and AgentOps
- Building evaluation datasets and running adk eval
- Logging, tracing, monitoring, and testing
- Deployment options: Vertex AI Agent Engine, Docker, and CI/CD
8. Agentic Security
- Prompt injection, insecure tool handling, and data leakage
- Mitigations: input sanitization, tool permissions, rate limiting, human review
- Model Armor guardrails and runtime threat detection
- OWASP Top 10 for LLM Applications review
Frequently Asked Questions
What does the AI Agents with Google's ADK Bootcamp: Build with Gemini, Protocols & AgentOps course cover?
Hands-on bootcamp. Build AI agents with Google's Agent Development Kit and Gemini, add tools and memory, use MCP and A2A, then evaluate and deploy. IT Dojo delivers it as live instructor-led training for government, DoD, and corporate teams.
How long is IT Dojo's AI Agents with Google's ADK Bootcamp: Build with Gemini, Protocols & AgentOps training?
Course length: 3 Half-Day Sessions (13.5 Hours). It is available as live remote online instruction or as a private session for your team.
Is this course available as live remote online training?
Yes. IT Dojo offers AI Agents with Google's ADK Bootcamp: Build with Gemini, Protocols & AgentOps as live remote online training led by an instructor in real time. Private sessions for your organization can also be scheduled.
Who should attend this course?
Python developers, AI/ML engineers, and technical leads or architects who want a structured introduction to building AI agents with Google's Agent Development Kit (ADK).
How do I register for this course?
IT Dojo training is employer sponsored. Your organization registers and pays for seats. Contact IT Dojo via the Request Training form or call 757-216-3656 to schedule this course for your team.