Agentic AI - Foundations

Day 1: Foundation - Intro do Coding Agents and AI Agents (March 4) - Claude Code and Gemini CLI - Using a coding agent in editor - Using coding agents Via CLI - Best Practices - Models - Context - Prompt Engineering - Structured Output - Introduction to Building AI Agents At the end of this class you should feel comfortable understanding how and when to apply agentic code in your development workflow. We will talk about tradeoffs and the different ways we can leverage these tools to produce better production code, while learning and also speeding up delivery. It is also a foundational class that will explain the primary concepts of building AI agents, where they work well and where they fail. We will build a few simple agents to explore these concepts. Day 2: Building Production Ready AI Agents + Advanced Coding Agents (March 11) - Smolagents and Pydantic AI - Agent Observability - Tools and MCP (Model Context Protocol) - RAG (Retrieval Augmented Generation) - Agent design patterns (workflow, orchestration, state management) - Advanced topics on Coding Agents (managing context, multi-agent, planning, Context driven development) We will cover the important concepts and build different agents exploring different use cases and agent architecture. We will discuss different AI agent patterns and why they matter. The goal is to understand the requirements and architectural design patterns to develop production-ready agents.