I design and build self-healing infrastructure — platforms that adapt, recover, and keep serving traffic when the underlying nodes don't cooperate. Kubernetes, GitOps, and chaos engineering, applied to production.
Building resilient systems means designing for failure, not around it — infrastructure that adapts and recovers on its own, so critical applications stay available under pressure, in dynamic, multi-cloud environments.
A Python/Click CLI for managing Kubernetes services end-to-end — cutting provisioning time from days to under 60 seconds. Automated governance policies and built-in chaos engineering (pod kills, CPU stress, network latency, AZ failure) keep deployments both fast and honest about their failure modes.
A multi-cluster Kubernetes platform spanning AKS and KinD, running the Payday SaaS app with full GitOps delivery. Hybrid cloud observability gives real-time metrics across Azure and local environments, so operations stay seamless regardless of where a workload lands.
Production fintech deployment on EKS, with a canary release strategy via Argo Rollouts and disciplined cost controls on the underlying cluster. Includes OIDC-based GitHub Actions pipelines and a hardened IAM setup for cluster auth.
A Flask + React application deployed on a kOps-managed AWS cluster, spread across multiple availability zones for high availability. Includes Kubernetes-native PostgreSQL manifests built to replace a managed RDS instance.
A multi-tenant SaaS platform for ops and team management, built on a hardened Node.js/Express API with Prisma/PostgreSQL and modular Terraform infrastructure. CI/CD pipelines run full security scanning and artifact signing end to end.