Why it matters
Versioning tracks what changed between document revisions. AI needs versions to cite the right era of a rule and to drop retired language from retrieval.
How it works
Store version IDs, diff summaries, effective dates, and bind chunks to specific versions. Migrate indexes on version bumps.
Example
Nintendo version history shows Refund Policy section 2 changed pro-rata rules on June 1. Chatbot answers after June 1 cite v9, not v8, with audit log proof.
Common mistakes
- 1Overwriting PDFs without version increment
- 2Chunks not tagged with source version hash
- 3No rollback path when a bad version ships
Your source changed. Your AI should know.
Track source freshness, detect changes, and review impact before answers drift.
Explore Source Maintenance