Building Graph RAG: A Beginner's Guide using C#, Neo4j and Microsoft Foundry

Discover Graph RAG with C# and Neo4j. Build a Knowledge Graph to improve AI accuracy and find deep connections. Here’s what you’ll learn: 🔥 📌 Why standard RAG struggles with summaries and complex relationships across large files. 📌 How to use an LLM to identify entities and relationships to build your graph automatically. 📌 How to combine Vector Search for "seed nodes" (a.k.a. the entry points) with weighted graph traversal using the reduce function. 📌 C# Implementation: A real world example using Neo4j and the Azure OpenAI SDK (to connect to models deployed in Microsoft Foundry) to build a complete pipeline. 📌 Different variations of Graph RAG including the ones which use Azure AI Search for retrieving the original data with citations. 👉 Blog post: https://deployedinazure.com/graph-rag... ----- 0:00 Intro 1:36 Quick App Demo 4:08 Classic RAG vs Graph RAG 9:23 How Graph RAG works 19:12 Building Knowledge Graphs using LLM 33:55 Graph Search: Vector Search + Graph Traversal 44:54 Other variations of Graph RAG 52:00 Summary ----- Follow me 👉 🌐 Blog: https://deployedinazure.com 💻 GitHub: https://github.com/deployed-in-azure/RAG 🔗 LinkedIn:   / deployed-in-azure