Search to production
Data
Takes a retrieval system from a local prototype through to production: the index model, the migration off the local store, historical backfill, live ingestion per source, and the instructions an agent needs to use it.
- Used when
- When a retrieval prototype works and the question is what it takes to run it for real.
- Loads first
-
- The local prototype skill first, then the production retrieval skill.
- The batch-workflow skill, for the backfill.
- Returns
-
- The prototype's status, command surface and verification results.
- The corpus, chunk, link, embedding, metadata and data-governance contract.
- The readiness gate result and the questions still open.
- The production index model and the migration plan off the local store.
- The backfill design, with cost, idempotency, metrics and replay.
- The ingestion design per source.
- The rollout plan, the verification, and the remaining limitations.
- Hands off
-
- Knowledge search — designing the retrieval agent itself