Hybrid Search (Dense + Sparse Retrieval)
Hybrid Search is an information retrieval technique that merges dense vector embeddings (semantic meaning) with sparse keyword search (exact match) using reciprocal rank fusion for superior accuracy.
How it works
Vector search alone can struggle with specific product SKUs, serial numbers, email addresses, or exact technical keywords, while keyword search alone misses synonyms and semantic meaning.
Hybrid search runs both retrieval algorithms in parallel and combines the result lists using algorithms like Reciprocal Rank Fusion (RRF).
The result is a balanced ranking that understands the conceptual intent of the query while respecting precise terminology.
How SiteMind uses it
SiteMind utilizes 768-dimensional embeddings alongside PostgreSQL full-text tsvector and trigram matching to achieve high-precision RAG retrieval before passing context to Gemini Flash-Lite.
Related terms
Semantic search
Semantic search finds content by meaning rather than exact keywords, so it can match a question to the right answer even when the words differ.
Vector database
A vector database stores content as embeddings and can instantly find the pieces whose meaning is closest to a query.
Retrieval-Augmented Generation (RAG)
Retrieval-Augmented Generation (RAG) is a method where an AI looks up relevant information from a trusted source before answering, so its replies are grounded in facts instead of guesses.