At Sonatype, I led a cross-region disaster recovery exercise for Maven Central and Nexus Repository, recovering the application stack with an RTO of about 2 hours 48 minutes against an 8-hour target. I also authored the team’s cross-region DR runbook.
I modernized Sonatype’s customer-facing download service, moving it from a six-system AWS stack to Cloudflare Workers and R2. The change reduced monthly infrastructure cost from about $3.5K to $15, with zero customer-visible downtime during DNS cutover.
Across reliability and platform operations at Sonatype, I defined SLOs and error budgets, helped sustain 99.9% uptime, and contributed to incident response that reduced MTTR by 35%. I also automated operational work, reducing toil by 50%.
Before Sonatype, I worked on DevOps at Infor and Ernst & Young GDS, improving CI/CD, Kubernetes operations, infrastructure automation, and monitoring. In a personal Google Cloud project, I provisioned a GKE-based service and added availability and latency SLOs with burn-rate alerts.
