Hybrid Search
Also known as: Hybrid retrieval, Sparse-dense retrieval
Combine BM25 (keywords) + dense embeddings (meaning); better recall than either alone.
A retrieval strategy combining lexical (BM25) and semantic (embedding) scores via reciprocal rank fusion or weighted sum. Captures both exact-keyword hits and conceptual matches that either method alone would miss.
In practice
The default production RAG setup post-2024. Interviews probe RRF, alpha tuning, and when pure dense suffices.
How it compares
Hybrid search is BM25 + dense embeddings combined; BM25 alone is the lexical half.
Related topics
Questions that mention this term
- Modern vector databases combine sparse (BM25) and dense vectors at the storage layer via two distinct patterns. Identify them.
- Architect search over any moment in a 100,000 hour video library
- Order the rollout of a semantic search relaunch over a legacy keyword engine
- RAG over a code repo can't find the exact function name a developer searches for: best fix?
- Spot the error in this hybrid search scoring code that combines BM25 and cosine scores.
- Match each retrieval technique to the specific failure it is designed to fix
Related terms
AI System Design
End-to-end design of production LLM systems: ingestion, retrieval, serving, eval, monitoring.
BM25
Classical TF-IDF-style lexical scoring: a surprisingly strong baseline, still the first stage of many hybrid RAG pipelines.
Chunking
Cutting documents into retrievable pieces before embedding them.
Context Engineering
Deciding what makes it into the model's context window (prompt + RAG + history + memory), and why.
Context Window
The max number of tokens a model can attend to at once.
Cosine Similarity
Compare two vectors by the cosine of their angle; the default similarity metric for embedding search.