- node_graphrag: multi-layer knowledge grounding across PostgreSQL

A high-performance Node.js engine for executing GraphRAG queries directly against PostgreSQL using multi-layer knowledge grounding.

This engine bridges the gap between structured relational data and unstructured LLM context by implementing multi-layer knowledge grounding. It leverages PostgreSQL (and the pgvector extension) to navigate complex entity relationships, transforming raw tables into a navigable knowledge graph. By executing local and global search patterns in a Node.js environment, it reduces latency for RAG applications that require deep reasoning across interconnected datasets.

https://github.com/vllm-project/node-graphrag