Crafting Your First AI Agent

Welcome to "Crafting Your First AI Agent", a hands-on lab designed to help you dive into the world of AI agents. Whether you're a developer, engineer, or just someone curious about how AI is evolving, this lab will walk you through the core concepts and give you the practical skills to start building your own AI agents.

Throughout this lab, you'll get hands-on with three popular frameworks for building AI agents today:

• LangChain — Great for building LLM-powered pipelines and retrieval-based systems like chatbots that can look up information.
• LangGraph — A powerful way to build stateful, memory-driven workflows, where agents can loop, make decisions, branch, and handle more complex processes.
• CrewAI — A framework built for multi-agent collaboration, where different agents work together based on assigned roles and responsibilities.

By the time you finish, you won't just know what AI agents are, you'll actually be able to build them, run them, and understand when to use simple pipelines versus multi-agent systems or dynamic workflows.

The goal of this lab is to introduce users to the core concepts of AI agents. By the end of the lab you should have a better understanding of what AI agents are, how they think, act, and observe the environment around them in order to accomplish a given task. Users will gain this understanding by taking a look at how agents can be created using three popular AI agent frameworks: LangChain, CrewAI, and LangGraph.

The lab walks the user through accomplishing the following:

1. Explanation of concepts, key terms, and technologies.
2. Create a basic agent using LangChain.
3. Create complex multi-agent workflow using LangGraph.
4. Create multi-agent architecture leveraging CrewAI.

This lab consists of a single VM with the following hardware and software:

• Windows 11 Jumpbox
• LangChain Framework
• LangGraph Framework
• CrewAI Agent Framework
• Nvidia NIM running Llama3.2-8B provided by the AI Proving Ground