CustomLabs
Retrieval

Vector Search (Semantic Search)

Also known as semantic search, similarity search

Vector search finds the nearest matches to a query by comparing embeddings, not exact keywords.

It lets a RAG pipeline surface a passage about "cancel my plan" for a user who typed "stop billing me."

In production it is usually paired with metadata filters and a reranker.

Raw nearest-neighbor results alone are rarely precise enough to hand straight to the model.

← Back to the full glossary

Source: https://customlabs.io/glossary/vector-search/

navigate select esc close