The next step in learning AI with .NET is knowing where it runs. Microsoft Foundry is the Azure layer around the model: identity instead of secret keys, deployments instead of hard-coded models, and one place to manage access. Here’s what it is and why it is step four.

The next step in learning AI with .NET isn't more code, it's understanding what's actually happening under the hood. Models, tokens, context, caching, temperature: five ideas that quietly control every AI bill and every AI response, made visible in this chapter.

In Chapter 1, AI helped you write code. In this one, your code talks to AI directly, and it turns out that's just an API integration you already know how to build. We wire Claude into a real .NET endpoint, wrap it in DI, and go parameter by parameter on what actually matters and why.

Every second LinkedIn post is about AI agents and RAG pipelines, and some quiet part of you is wondering if you're falling behind. But you're not starting from zero. Here's where an experienced .NET engineer actually starts with AI.