GenAI is undoubtedly becoming increasingly important. However, building intelligent applications and integrating AI into software is not an easy task. Real-life solutions have numerous requirements that go beyond the scope of an experimental application built for Data Science purposes or simply playing around with prompts.

In this demo-driven session, you will learn how to build AI agents integrated into applications of any type. You will see many demos in .NET/C# that demonstrate building GenAI applications capable of running almost anywhere and on any OS. Even if C# is not your primary language, feel invited to learn about concepts like Agents and Plugins, which will help you build your custom copilots.

We will start with core topics such as tokens, completions, cosine similarity, embeddings, and different model types and then combine them to build custom copilots using Azure OpenAI, Azure Open AI and Semantic Kernel. Once you understand how everything works, you will be able to teach existing models to perform tasks they usually cannot do.

This session is a technical presentation designed for developers, architects, and technical decision-makers.