Articles for people who
ship the stack
Rewrites on React, Node.js, TypeScript and AI. A translated body is shown when it exists. If it does not, you see the English text and a note that says so.
Tagged: Hybrid Search
Let Exact Terms and Meaning Work Together in Search
Mix BM25 and semantic search candidates without treating their scores as comparable: fuse by rank, dedupe by identity, and give each retriever its own input.
2 min readHybrid Agent Memory: Fusing BM25 and Vector Search With RRF in Python
Learn why pure vector search fails as agent memory, how Reciprocal Rank Fusion merges BM25 and dense results in Python, and when GraphRAG summaries help.
7 min readBeyond Top-K: Relevance Thresholds, Hybrid Search and Reranking in RAG
See why a vector database plus an LLM is not a production RAG system, and how chunking, similarity thresholds, hybrid search, reranking and evaluation close the gap.
7 min readReducing Hallucinations in a Medical RAG Chatbot Pipeline
Cutting hallucinations in a medical RAG chatbot with hybrid search, reranking and a strict refuse-to-fabricate policy, after plain vector search fell short.
9 min read
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.