Platforms, not ticket queues
The best platform team is the one product engineers rarely need to talk to. I build paved roads — golden Terraform modules, self-service APIs, sane defaults — so teams ship without waiting on me.
Senior Platform & DevOps Engineer based in Gurugram, India — building cloud platforms that engineering teams actually enjoy using.
Platform and DevOps engineer with 7+ years building scalable, highly available cloud infrastructure across AWS, GCP and Azure. I design self-service capabilities that let engineering teams ship without filing tickets — Infrastructure as Code with Terraform, hardened Kubernetes platforms, DevSecOps guardrails and CI/CD that people actually trust. I have run mission-critical systems for a regulated bank, driven FinOps cost programmes, and led DevOps teams across concurrent client engagements.
In practice that means I spend my time on three things: making infrastructure reproducible, making delivery safe, and making both of those someone else's self-service instead of my inbox. I have run production workloads on AWS, GCP and Azure — including EKS, GKE and AKS clusters — wrapped in Terraform, fronted by Kong and Istio, and delivered through pipelines in Azure DevOps, Cloud Build, CodePipeline and Jenkins.
Most recently at Kotak Mahindra Bank, that work happens under real regulatory pressure: zero-downtime releases on customer-facing banking systems, live CVE visibility across the container estate, and automated database backups into S3 that satisfy both compliance retention and disaster recovery. Before that, as Lead DevOps Engineer at FiftyFive Technologies, I led a team delivering multi-cloud platforms for several clients at once, and wrote the FastAPI service that let their developers manage Kubernetes resources themselves.
I hold six certifications — including CKA, CKS and CKAD — and I write about the things I break and fix on the blog. The tools section is the practical side of that: small, fast, browser-only utilities I got tired of looking up.
The best platform team is the one product engineers rarely need to talk to. I build paved roads — golden Terraform modules, self-service APIs, sane defaults — so teams ship without waiting on me.
CKS-certified and coming from a regulated bank: guardrails belong in the pipeline and the cluster, not in a review meeting at the end. Visible CVE posture beats a quarterly PDF scan report.
Zero-downtime Blue-Green releases, tested backups, alerting that fires before customers notice. Reliability work is unglamorous and it is the whole job.
FinOps is not a spreadsheet exercise. Right-sizing, lifecycle policies and killing idle resources are design decisions that belong next to latency and availability.