A hands-on walkthrough of building a multi-agent AI application using Google's Agent Development Kit (ADK) with A2A protocol support, powered by Gemini Flash LLM. The system consists of five specialized agents (Researcher, Judge, Orchestrator, Content Builder, Course Builder) built in Python, tested locally with Gemini CLI, and deployed to Azure AKS. The guide covers environment setup with pyenv and nvm, local development and debugging via Makefile targets, Kubernetes deployment, and using Gemini CLI for code review and log analysis.
Table of contents
Building a Multi-Agent System | Google CodelabsGitHub — pyenv/pyenv: Simple Python version managementGitHub — nvm-sh/nvm: Node Version Manager — POSIX-compliant bash script to manage multiple active…Agent Development Kit (ADK)Agent Development Kit (ADK)Get xbill ’s stories in your inboxMulti-Agent A2A with the Agent Development Kit(ADK), Cloud Run, and Gemini CLI2.5K Impressions