GenAI for .NET and Azure Developers: The Ideas in C#, Plus 3 Portfolio Projects

Most GenAI material is written in Python. If you work in .NET and Azure, this guide gives you the same ideas in your language, with the Microsoft library and the Azure service for each one.

Part 1, the ideas in C#: calling a model, structured output and tools, RAG on Azure AI Search with permissions, agents with Microsoft Agent Framework and MCP, evals in .NET, and a small production setup on Azure.

Part 2, three projects with full source code:

• PolicyBot: document question answering where each person only sees what they're allowed to see

• HelpdeskAgent: an agent whose actions pass through a policy gate in code, with human approval

• EvalGate: an eval set and a release gate that fails the build when quality drops

Part 3, the interview: resume lines, a one-minute story for each project, and short answers to 14 questions .NET developers get asked.

About the code:

• C# on .NET 10, with 46 automated tests

• No NuGet packages, no API key and no Azure subscription needed to build, run or test

• Runs offline with small stand-ins for the model; set four environment variables to call Azure OpenAI

• Pipeline files for GitHub Actions and Azure Pipelines

Plain facts: the Azure OpenAI client was tested with canned responses, not against a live Azure resource. Samples that use Microsoft's libraries follow Microsoft's documentation as of November 2026 and were not compiled for this guide. Both are marked clearly inside.

What you get:

• The guide, 41 pages, as a dark screen edition and a light print edition

• GenAI-DotNet-Projects.zip with all source code

Refunds: because this is an instant download, all sales are final. If a file won't open or the download fails, email illustratedengineer@kronosautomation.com and we'll fix it.