Show HN: Replaced Neo4j with pure vector search for Graph RAG
Category: ai-ml
Tags: rag, vector-search, knowledge-graph, llm, multi-hop-qa
Score: 7.3/10 (Innovation: 7, Technical: 7, Documentation: 8, Utility: 7)
Vector Graph RAG is a Python library that implements Graph Retrieval-Augmented Generation using pure vector search in Milvus, eliminating the need for a traditional graph database like Neo4j. It's interesting because it claims to achieve state-of-the-art multi-hop question answering performance with a simpler, single-pass LLM reranking architecture, significantly reducing operational and computational costs compared to iterative agent-based approaches.
Target audience: data engineers, ai engineers, backend devs
Repository: https://github.com/zilliztech/vector-graph-rag · Python · MIT · 22 stars
View on Hacker News