ai.rag

ai.rag

Point-in-time retrieval-augmented generation workflows.

Functions

Name Description
ask Retrieve point-in-time context and generate a source-grounded answer.
index Chunk and index source documents with stable content-derived IDs.

ask

ai.rag.ask(question, *, knowledge, client, model, as_of, where=None, limit=8)

Retrieve point-in-time context and generate a source-grounded answer.

index

ai.rag.index(documents, *, store, chunk_size=512, chunk_overlap=64)

Chunk and index source documents with stable content-derived IDs.

Back to top