Every .NET team is being asked the same question: “How can we implement LLMs in our application?”.

In this hands-on workshop, you’ll learn how to build production-ready AI features in C# using Microsoft.Extensions.AI, an exciting new provider-agnostic library for modern .NET AI development.

We’ll start using free local models via Ollama to build an AI chat client using the IChatClient interface. From there, we’ll dive into the software engineering best-practices that make AI features maintainable: unit testing, and evaluating natural language responses.

On Day 2, we'll add image generation using IImageGenerator, covering unit testing and guard-rails for AI-generated images.

Then we’ll build a Retrieval-Augmented Generation ("RAG") pipeline using IEmbeddingGenerator and vector stores and use it to ingest proprietary data in our app to train + augment our models.

Finally, we'll explore how the same application can move easily between local models, Azure AI Foundry, AWS Bedrock, on-device AI, and more, without changing the core business logic.

By the end of the workshop, you will have built a working AI-powered .NET application and, more importantly, will understand the architectural patterns needed to responsibly and reliably bring AI into existing .NET applications.

This is not a prompt engineering workshop. This is real-world AI engineering for .NET developers.