I don't think of DevOps as a toolchain. I think of it as a feedback loop — the shorter the loop between writing code and running it reliably in production, the better the engineering culture.

Most of my work lives in that gap:

• Infrastructure that's consistent and auditable, not click-ops and hope
• Pipelines that catch problems early, not after the on-call gets paged
• Observability that tells you why something broke, not just that it did
• Access patterns that follow least-privilege by design, not by afterthought

"The goal isn't automation for its own sake. It's reducing the cognitive load on engineers so they can focus on what actually matters."

current_role:    Lead DevOps Engineer (Kavia AI)
primary_focus:
  - Scaling EKS environments for AI workloads
  - Hardening CI/CD pipelines for safer, faster releases
  - Improving Terraform module structure and remote state management
exploring:
  - GitOps patterns with ArgoCD
  - eBPF-based observability tooling
  - Cost attribution and FinOps practices on AWS
open_to:
  - Platform engineering collaborations
  - DevOps consulting on AWS and Kubernetes
  - Open source contributions in the cloud-native space

These are the tools I reach for daily and the ones I've built production systems with:

Cloud & Platform

Delivery & Automation

Observability

Not a portfolio dump — just a few problems I found genuinely interesting to solve:

🏗️ Multi-AZ EKS from scratch with Terraform
Built the full stack: VPC, subnets, OIDC federation, managed node groups, autoscaling, and add-ons. The goal was a platform another team could onboard onto without needing to understand the underlying AWS plumbing.

⚡ Cut deployment cycle time by 40%
Replaced manual release steps with end-to-end Jenkins + GitHub Actions pipelines — Docker multi-stage builds, ECR image scanning, Helm rolling deploys, and automated rollback gates. Less toil, fewer incidents.

🔍 Observability stack that teams actually used
Deployed Prometheus + Grafana + Loki on Kubernetes. The difference wasn't the tools — it was building dashboards around questions engineers actually asked, not metrics that were just available.

🔒 IAM patterns that don't break things six months later
Designed least-privilege role structures with OIDC service accounts, audit logging via CloudTrail, and Security Hub compliance workflows. The real win was making it easy for developers to request access correctly, not just saying no.

• 🌱 Started as a network engineer — that foundation shapes how I think about security, latency, and failure modes in ways that pure cloud engineers sometimes miss
• 🧱 I prefer boring, reliable infrastructure over clever infrastructure. Clever is a liability at 2am
• 📖 I write notes on problems I've debugged — it's the fastest way I know to learn something properly
• 📍 Based in Ahmedabad, India — open to remote-first roles globally

If something here resonates, feel free to reach out — always happy to talk platform engineering, Kubernetes war stories, or AWS cost surprises.