Skip to Content

Core Concepts

OKF retrieval vs vector RAG

Which questions each kind of retrieval answers best, and when to link both to one engine.

Vector RAG and Docklee OKF retrieval are good at different questions. Vector RAG matches meaning across loose text. OKF retrieval follows how things connect. Most real knowledge benefits from both, and one engine can read both at once.

Where each one wins

QuestionBetter withWhy
"How does A relate to B?"OKFIt follows typed links.
"Who is affected if A changes?"OKFIt walks several links in one search.
"Which tables hold customer data?"OKFIt filters by concept type.
"What's the current policy?"OKFIt drops superseded concepts.
"Find anything like this paragraph."VectorIt matches by meaning across loose text.
"What did the manual say about X?" on long, unstructured textVectorLong prose doesn't break into clear concepts.

When content has no real structure, vector search is the simpler choice, and OKF adds little. OKF pulls ahead as soon as relationships, types, or freshness matter.

Side by side

OKFVector
StoresConcepts and typed linksChunks and embeddings
DimensionNoneOne per container
Ingest costHigher. Building concepts and links takes model calls.Lower
Query costNear zero for structural questionsOne embedding per question
Shows its reasoningYes, as a path of linksNo
PortableYes, as an open bundle you can exportNo

Use both

Link vector sources and OKF sources to the same engine, and every question searches both. The results are merged into one ranked list. See Vector and OKF together.

A common split:

ContentContainer
Help center articles and manualsVector
Data catalog, policies, org structure, product catalogOKF

Next steps

You know where OKF fits. Turn your content into concepts next.

On this page