Vector Embedding
A numeric representation of text where passages with similar meaning sit close together, which is what lets a search match intent rather than exact keywords.
Why keyword search fails on technical documents
A buyer asking about 'rust resistance in salt water' will never keyword-match a specification that says 'pitting resistance in chloride service'. Embeddings place both near each other, so the right clause is retrieved.
Chunking decides quality
Documents are split before embedding, and bad splitting is the most common cause of poor retrieval — a tolerance table cut in half retrieves as nonsense. We chunk on document structure, not on a fixed character count.
Related terms
Where this applies
StancorIQ
Need help applying this?
Stancor Technologies engineers answer specification questions directly. Contact us.