Skip to content

RAG Knowledge Base Management

Description

Provides tooling to ingest, chunk, embed, and index enterprise content — documents, web pages, FAQs, product manuals — into a vector knowledge base. Includes refresh pipelines to keep the knowledge base current as source documents change.

Canonical use case

A software vendor ingests its entire documentation portal into a RAG knowledge base, enabling the support bot to answer version-specific questions with cited passages from the official docs.

Open Items

  • [ ] Canon alignment — populate canon_axiom_refs or confirm no existing axiom applies
  • [ ] Dependency assessment — set dependencies_assessed: true once SA has reviewed the full chain
  • [ ] effort_estimate — replace 0 with rough engineering days (order of magnitude)
  • [ ] public_description — write the public-facing description before publishing