Context for agents
Bring both signals into the context your agent works with.
DATABASE / HYBRIDThe right words.
The right words.
The right meaning.
Combine full-text search with vector similarity so your agents and applications can use both exact terms and contextual relevance.
“Lightweight laptop backpack for commuting”
FULL TEXT
Words
laptop · backpack
VECTOR
Meaning
portable · everyday carry
COMBINED CONTEXT
Everyday backpack
A result informed by both the words and the intent of the request.
Illustrative retrieval flow
Keyword precision
Keep names, phrases, and specific terms in the retrieval process.
Semantic relevance
Find related content even when it uses different language.
Use both sides
of the question.
Start with the request
Take the agent’s retrieval task or the user’s question.
Search words and meaning
Use full-text and vector search as complementary retrieval signals.
Return useful context
Bring relevant records back to the agent or app.