All services experienced a brief interruption (~10 minutes) during a code deployment. Updates included odds pipeline fixes, manager intelligence, and API improvements. All services auto-recovered. No data loss.