I built an open-source Knowledge Graph pipeline with hybrid retrieval to improve LLM multi-hop reasoning [P]

Hey everyone, I built an open-source full-stack pipeline (Django + React) that constructs a Knowledge Graph from raw text, detects thematic communities, and uses hybrid search to solve the "lost in the middle" problem in standard vector retrieval. The Pipeline: Ingestion & Chunking: Raw text is cleaned, parsed, and split into overlapping chunks to preserve local context. Graph Construction: spaCy extracts named entities from each chunk. A weighted co-occurrence graph is built using NetworkX , mapping which