Enhancing RAG with Neo4j Knowledge Graph
summary
Claims that coupling Neo4j graph DB with vector+graph retrieval lets RAG answer cross‑document and “global” queries. A multilingual news graph is built by crawling with Fundus, chunking and embedding paragraphs (Gina AI), extracting entities with Gliner, storing nodes/relations in Neo4j, and querying via LangChain either by vector search + graph‑reversal or by LLM‑generated Cypher. Uses Snowflake Arctic LLM; demo produces summaries and source/date counts but shows retrieval limits, occasional NER/model errors, and live demo network failures.
AI-written summary · about
chapters (7)
- 0:00Introduction and agenda
- 2:46LLM limitations and RAG
- 7:56Knowledge graphs explained
- 11:08Techniques for RAG with graphs
- 15:17System implementation
- 20:25Live demo and walkthrough
- 32:25Q&A and closing
related talks
30mAgentic RAG with Knowledge Graphs Explained
16mEnhancing Retrieval-Augmented Generation with GraphRAG Patterns in Neo4j
30mGraph RAG: Bringing together graph and vector search to empower retrieval
19mBeyond Retrieval: Adding a Reasoning Layer to RAG
25mEnhancing Job Matching With Knowledge Graphs and RAG
36mGraphRAG Explained 🔥 Building Smarter LLMs
25mAdaptive GraphRAG: A Framework for Knowledge Graph Quality, Consistency, Evolution
01h 17mGraph Database and Retrieval Augmented Generation