Articles for people who
ship the stack
Original rewrites on React, Node.js, TypeScript and AI — practical notes from the same engineering practice behind our operator software. Article bodies are in English.
Tagged: bm25
Hybrid Search for Enterprise RAG: Dense Alone Is Not Enough
Exact ids and rare codes break embeddings; BM25 plus dense fusion in parallel covers the real query mix.
2486 wordsRead articleRAG, vectorless RAG, and GraphRAG compared
Classic vector RAG, lexical vectorless retrieval, and GraphRAG: chunking, embeddings, BM25, multi-hop graphs, and when each approach earns its keep.
1834 wordsRead articleLet Exact Terms and Meaning Work Together in Search
Combine lexical and semantic candidate lists without treating incompatible search scores as interchangeable.
572 wordsRead articleHybrid RAG Retrieval with pgvector, BM25 and a Cross-Encoder Reranker
Learn why pure vector search misses part numbers and error codes, and how to combine pgvector, BM25 and reranking in LangChain for precise RAG retrieval.
1524 wordsRead article
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.