Loading video player...
Build production-ready multi-agent AI systems with Google's Agent Development Kit (ADK) in this complete 3-hour crash course. We'll go from zero to a fully deployed WealthPilot AI Wealth Advisor ā covering every core ADK primitive along the way: agents, tools, callbacks, session state, memory, artifacts, code execution, planning, parallel/loop agents, the runner, and local Docker deployment with a Next.js frontend. By the end of this video, you'll have a complete, working multi-agent system that analyzes stocks, builds investment portfolios, and generates downloadable PDF reports ā deployed locally with Docker Compose, and Google Cloud Run š Chapters: 0:00 ā Intro: Welcome & What We'll Build 3:12 ā The Problem with Monolithic LLMs 7:53 ā What is Google ADK? 16:52 ā Python Dev Environment Setup 21:37 ā Go / Java / TypeScript Setup 23:39 ā ADK Core Components Overview 36:46 ā Building Your First Multi-Agent System (Agents) 56:11 ā ADK Tools: Custom Tools with yFinance 1:05:44 ā ADK Callbacks: Guardrails & Validation 1:21:13 ā Session, State & Memory 1:48:17 ā Code Execution & Financial Calculations 2:00:30 ā Artifacts: Saving Portfolio Reports (PDF) 2:15:58 ā Planning with Thinking Budget 2:20:17 ā Parallel & Loop Agents 2:31:23 ā ADK Runner Deep Dive 2:42:12 ā Deploying ADK Locally (Docker + Next.js) š Want more? The full course also covers: ⢠Deploying ADK to Production with Google Cloud Run ⢠Deploying to Vertex AI Agent Engine (the fast path to production) ⢠ADK Observability: structured logging, distributed tracing & Cloud Trace ⢠Production monitoring with AgentOps (session replay, analytics dashboard) š Full Course on Udemy: https://www.udemy.com/course/building-multi-agent-systems-google-adk-masterclass/ š¦ Follow me on X: https://x.com/_thinking_reed GoogleADK #AIAgents #MultiAgentSystems #Python #LLM #AgentDevelopmentKit