Embedding Pipeline Service  ·  View 13 of 22  ·  Runtime

Critical Flow — One Edit Becomes Retrievable

Seventeen messages from a saved document to a retrievable version, including the point at which the new version becomes visible at all.

Editable source SVG draw.io All views
Document service Ingest + admission Change log Extract + chunk Chunk ledger Embedding fleet Vector index Retrieval API 1. changed(doc, v41) 2. quota + lane: interactive 3. append, key = doc 4. durable 5. 202 accepted 6. deliver, in order per doc 7. version 40 known? 8. 41 chunk hashes 9. extract, normalise, chunk 10. diff: 3 new, 38 reused 11. embed 3 chunks 12. cache miss, batch, infer 13. upsert 3 vectors 14. mark v41 complete 15. flip visible version 16. search (contract-scoped) 17. v41 chunks + staleness Critical Flow — One Edit Becomes Retrievable Message 15 is the atomicity point: until all three vectors land, retrieval keeps returning version 40 rather than a half-updated document. v 1.0 · owner Data & AI Platform Architecture · date 2026-10

The atomicity point

  • Message 15 is the architecture: the new version becomes visible only once every chunk of it is indexed. Until then retrieval keeps returning the previous version.
  • A half-reindexed document is strictly worse than a stale one — it matches on old wording and new wording at once, and no reader can tell which they are looking at.
  • The ingest 202 at message 5 is returned on durability in the change log, not on being indexed. The accept path never waits on a GPU.

Where the cheapness comes from

  • Messages 7 to 10: the ledger is asked what the previous version's chunks were, and the diff reduces 41 chunks to 3.
  • Message 12: those 3 are looked up in the vector cache first. On a reverted edit all 3 are hits and no inference happens.
  • Delivery from the change log is in order per document (keyed partitioning), so two rapid edits cannot race into an older version winning.

Assumptions

  • 41 chunks is around the p75 document; the mean is 12 and p99 is 310.
  • Interactive lane p50 30 s end to end for this flow, p99 10 min (stated assumption).