
GitHub
Platform engineer who builds CI/CD systems that make deploy days boring. Power of 2.

Built zero-downtime pipeline going from 2 deploys/week to 15+/day
Height deployed twice a week with 15-minute maintenance windows. Each deploy required a manual checklist of 23 steps and the on-call engineer's presence. Deploys failed 18% of the time, and rollbacks took 45 minutes. The team was growing from 8 to 25 engineers and the process wouldn't scale.
Built a fully automated CI/CD pipeline using GitHub Actions + Kubernetes blue-green deployments. Implemented automated canary analysis with Datadog metrics gates, database migration safety checks, and one-click rollback. Added preview environments for every PR using ephemeral Kubernetes namespaces.
Went from 2 deploys/week to 15+ deploys/day. Deploy failure rate dropped from 18% to 0.8%. Rollback time went from 45 minutes to 90 seconds. Eliminated maintenance windows entirely. Engineering velocity (PRs merged/week) increased 3x.

QA testing happened on a shared staging environment. Only one feature branch could be tested at a time, creating a bottleneck for the 25-person engineering team.
Built ephemeral preview environments using Kubernetes namespaces + Terraform. Each PR automatically gets a full-stack environment with seeded test data.
QA throughput increased 5x. Bug escape rate to production dropped 60%. Engineers stopped fighting over staging.