Copilot, Cursor, and Custom LLMs: Navigating the New .NET Developer Experience

It feels like every two weeks there’s a new "game-changing" AI tool that promises to write our code for us. But for those of us working in the .NET ecosystem, the reality is a bit messier. We’re caught between the tools we know (Visual Studio), the new AI-native editors everyone is talking about (Cursor), and the reality that public LLMs often struggle with our internal business logic and private libraries.

I’ve spent the last year trying to figure out where the "productivity" ends and the "hype" begins. In this session, we’re skipping the "Hello World" demos and getting into the actual friction of the new .NET developer loop.

We’ll look at:

The Tooling Split: When should you stick with the deep integration of Copilot in Visual Studio, and when is it actually worth switching to Cursor for its "Composer" features?

Context is Everything: How to stop the "hallucinations" by giving these tools better context of your specific .NET solution, including using local LLMs for the stuff you can’t send to the cloud.

The "Senior Dev" Reality Check: How our roles are changing from just writing lines of C# to auditing AI-generated PRs and managing "agentic" workflows without breaking the build.

This isn't a sales pitch for any specific tool. It’s a practical look at how to build a modern .NET development environment that actually helps you ship faster, rather than just adding another subscription to your credit card.

What you’ll learn:

A side-by-side look at Copilot vs. Cursor on a real C# codebase.

How to use RAG and local models (like Ollama) to talk to your own docs and legacy code.

Practical tips for "Context Engineering" to get better C# refactoring results.