Cut vulnerability remediation from 30 days to 5.
The team was running a quarterly-patch culture: a scanner finds something, a ticket sits, a patch window eventually happens, everyone hopes. I replaced it with an event-driven pipeline that triages, assigns and verifies, with proof of remediation built in rather than assumed.
- Normalised findings from multiple scanners into a single queue.
- Enriched each finding with asset owner, criticality and exploit-in-the-wild status.
- Auto-ticketed with an SLA per severity; isolated actively exploited CVEs.
- Re-scanned on patch deployment and reopened automatically on regression.