Hybrid Search
Combining keyword search and vector search, then fusing the two result sets.
In plain language
Vector search captures semantic similarity but misses exact IDs, proper nouns and rare terms; keyword search such as BM25 is the opposite. Hybrid search runs both and fuses results with reciprocal rank fusion or weighted scores, covering both semantic and exact-match needs. Enterprise knowledge bases, legal and medical search benefit most. The hard parts are score normalization and weight tuning, which require a real question set.
Example
For “X200 battery life,” keyword search matches the model number exactly while vector search handles “battery life/usage time.”
Recommended tools
Related terms
Embedding
Converting text, images or other content into vectors whose similarity can be computed.
Reranking
Re-scoring and reordering retrieved passages with a stronger model after initial retrieval.
Retrieval-Augmented Generation
Retrieve relevant passages from a knowledge base, then have the model answer from them.
Agent Memory
A mechanism for an agent to store and retrieve user preferences, past tasks and key facts across sessions.