I've spent the last ten years in the space between application code and production, the messy, interesting place where deployments either feel invisible or keep you up at night.
I got into DevOps and platform engineering because I liked making that gap smaller. Fewer manual steps. Fewer surprises on release day. More time for teams to focus on what they're actually building. That instinct, to make things reliable and repeatable, still drives most of what I do.
A lot of my work has started with pipelines. Not because CI/CD is glamorous, but because it's where trust gets built. I've designed application and infrastructure pipelines across Jenkins, GitHub Actions, ArgoCD, and Terraform, from designing environments for Python and Node services to full AWS environments provisioned as code. When a team believes in their path to production, they ship with confidence. That's the bar I hold myself to.
At enterprise scale, Kubernetes stops being a single cluster and becomes a shared foundation: many teams, many tenants, one platform. I've designed and operated multi-tenant EKS environments with proper isolation, RBAC, and guardrails across hundreds of clusters. The hard part isn't the manifests. It's making the platform feel safe and predictable for everyone on it, including whoever is on-call at 2 a.m.
That experience led me to internal developer platforms. I led the architecture of an IDP that moved teams off brittle VM workflows onto Kubernetes, with self-service for developers and consistent standards for the organization. Reusable Terraform modules, GitOps with ArgoCD, sensible paths for Python and Node applications. The goal was straightforward: make the right thing the easy thing.
Cloud cost is where good engineering meets real business impact. I've optimized over $250K in lifetime cloud spend through right-sizing, event-driven automation, and treating cost as a metric you watch, not a surprise you explain. Reliability and efficiency usually turn out to be the same conversation.
None of it holds together without observability and governance. I've built monitoring stacks with Prometheus, Grafana, Loki, and Datadog, and embedded security and compliance checks into CI/CD so problems surface early, not in production. A platform that runs fast but fails quietly isn't a platform I'd want to operate.
If you're reading this, you're probably deciding whether we'd work well together. I care about that. I enjoy mentoring engineers, partnering with product teams, and solving problems that sit at the intersection of people and infrastructure. However you found your way here, whether hiring, collaborating, or just curious, I'm glad you stopped by.