Combine retrieval signals
Bring results from term matching and vector similarity into one ranked candidate set.
Desired outcome
Produce a candidate set informed by more than one retrieval signal.
Explore from the outcome
Where this function can lead
These documented connections are starting points for selection, not a ranked list.
Ways to do it
Methods that can help achieve this outcome.
- Hybrid text and vector retrieval
Combine lexical matching with vector similarity and fuse or rerank candidates.
Link evidence: pgvector project documentation
Typical inputs
- Text query, indexed text and vector representations
Typical outputs
- Combined and ranked candidate documents
Sources
Last reviewed 2026-09-30