We'll start by breaking down what actually makes an AI agent different from a standard LLM call, the reasoning loop, tool use, and decision-making that turn a model into something that can act, not just respond. Just enough theory to get everyone on the same page.

Then we go hands-on. We'll build a real agent together using LangChain, and trace what's happening under the hood using either Azure AI Foundry or LangSmith.

By the end, you'll have built your first working agent and understand the core pieces you need to build your own from scratch.