I built Enterprise RAG web service and the corresponding Desktop Client, all in ASP.NET Core 8/C# using Azure OpenAI API and Azure Cognitive Services and Azure Cosmos vector DB,
I read somewhere that neo4j knowledge graphs can significantly improve the accuracy of RAG.
I know just a bit about neo4J and am drowning in the ocean of data and topics. What is the recommendation for "where do I start" in order to be able to build graphs from the private corpus of data that I ingest into Azure Cosmos vector DB and integrate it with my RAG web service?