AI DevCamp

# Intro to Agents & Codelabs

Renuka Kelkar

Renuka leads a focused 2-hour hands-on workshop introducing AI agents — how they think, plan, and act. You'll work through 14 guided codelabs that take you from agent fundamentals to practical implementations. By the end of the day you'll have a working understanding of agent loops, tool use, and how to wire them together.

# MCP · Advanced Agents · Deployment

Renuka Kelkar

Renuka returns for a focused 1-hour session diving into the Model Context Protocol (MCP), advanced agent patterns, and how to deploy your agent to the cloud. We then open the floor for codelab questions and live troubleshooting — bring your blockers!

# Google ADK & Vertex AI

Saoussen Chaabnia

Saoussen Chaabnia leads a deep 2-hour build session: Full-Stack Multi-Agent App with Google ADK & Vertex AI. You'll go from zero to a deployed multi-agent application backed by Google's Agent Development Kit and powered by Vertex AI — the same stack used in production at scale.

# Assignment Completion & Project Showcase

The final stretch before the closing ceremony. We review assignment completion, go through codelab 3, and participants showcase their projects to the group. Get feedback, celebrate your work, and prepare for the in-person closing.

# Deep dive into Agent-to-Agent protocol

Nishi Ajmera

Agent-to-Agent (A2A) protocols define how independent AI agents discover each other, exchange structured messages, and collaborate on tasks across system and organizational boundaries. In this talk, we'll go beyond the basics and explore how agents communicate using standardized primitives like Agent Cards, tasks, messages, and artifacts. We'll look at how agents delegate work, negotiate capabilities, and coordinate execution. We'll also cover real-world design patterns for building interoperable, multi-agent systems where agents built on different frameworks can seamlessly work together using a common protocol layer.