Cloud & DevOps Engineering

Cloud work is successful when shipping is routine, incidents are understandable, and the bill does not surprise anyone. I build delivery paths around repeatable deploys, clear ownership, sensible defaults, and enough visibility to answer what changed, what failed, and what it costs.

The failure modes are familiar: a container works locally but not in CI, a serverless function starts cold without its dependencies, a deployment succeeds while routing is broken, or one noisy workload starves everything else. I make those boundaries explicit with reproducible images, automated pipelines, health checks, logs, least-privilege access, and resource requests and limits. The aim is not elaborate infrastructure—it is a system the team can confidently change on a Friday.