Getting Started with Retrieval Augmented Generation (RAG)

Retrieval Augmented Generation (RAG) seems to be the most popular way people are using AI in their own applications. If you have any business applications that could be enhanced by adding the ability to search your data (database or documents) using a ChatGPT like interaction, then RAG may be the starting point for you.
In this session, I will do a quick introduction of the parts of a RAG implementation - though I do assume you are already familiar with what Large Language Models (LLM's) are and how to use them. The majority of the session will be spent looking at different ways to implement the RAG pattern using .NET with Azure and OpenAI services. We'll start with a simple naïve implementation and get into more detail on ways to improve the functionality.
Demos include:

1. A simple local Blazor web application using SQL Server and Azure Open AI to understand RAG, vectors/embeddings
2. A full featured RAG application with a Blazor front-end and Azure AI Search for storing vectors for searching unstructured data (ie. Pdf files)
3. An additional demo that shows RAG with structured data (semantic search on data from a database) also written in .NET

Watch live on YouTube if you are not able to be there: https://www.youtube.com/watch?v=JMX-qyRuAzE

Location Sponsor: Concurrency
Food Sponsor: TEKSystems