Semantic Search / Embeddings

Retrieval works by meaning rather than keyword, so a query finds the passage that answers it even when it shares no words with it. Embeddings are generated and stored inside your deployment, and access is scoped per user group. Embeddings can be regenerated when the model changes, without touching the source documents.

You can use the models we operate or bring your own model and we run it on your cluster. Work arrives as an API call, a batch over a folder or bucket, or a resident on-premise service, and results return as structured output with a per-job audit record.