GraphRAG

Sep 2025GenAI / RAG

Summary

A GraphRAG pipeline that reads table and column schemas from Databricks SQL and builds a knowledge graph of tables, columns, and their relationships in Neo4j, giving LLMs richer context for answering questions about the data.

Highlights

  • Extracts schemas from Databricks SQL tables
  • Builds a Neo4j knowledge graph of tables, columns, and relationships
  • Graph context for retrieval-augmented LLM question answering
CategoryGraphRAG / Knowledge Graphs
StackPython, Neo4j, LangChain, Databricks SQL

Want to talk about this project?

I'm always happy to chat about agents, RAG, and GenAI engineering.

Get in touch