Platform & reliability engineer. I make Kubernetes systems faster, compliant, and dependable.

Ex-Oracle (cluster infrastructure) · IIIT Hyderabad (B.Tech + MS by Research) · Outreachy alum (TensorFlow)

• Cluster controller ownership. Owned controller logic for a distributed Kubernetes platform (~20 pods, multiple replicas). Designed key assignment that guarantees mutual exclusion across concurrent controller replicas.
• 21s faster scale-up. Reworked scheduling and synchronization logic to cut cluster scale-up latency by 21 seconds against production SLO baselines.
• FIPS 140-2 DNS, zero downtime. Shipped a FIPS-compliant DNS service into the production Kubernetes networking stack with a zero-downtime cutover, using plugin chaining. Adopted by 2 internal teams.
• Operations. Introduced structured log levels to cut production log noise; on-call rotation and incident postmortems with SRE and platform teams.

A webhook relay and delivery platform, built to practice production DevOps and SRE end to end. Repos: ferrum-webhook-relay-core · infra

Built so far

• FastAPI gateway + async worker over a Redis queue, PostgreSQL for persistence
• Kubernetes deployment with HPA, readiness/liveness probes, and resource limits
• Prometheus metrics for throughput, p95 queue latency, and processing rate
• GitHub Actions → GHCR → cluster deploys; Docker Compose for local parity

Measured

• 135 req/s at 0% errors (k6, 50 VUs, 60s)
• Client acknowledgement under 5ms, independent of downstream endpoint latency
• Cache-aside on Redis: lookups down from 22ms to 1.2ms on cache hits

Next (roadmap, not done yet)

• AWS deployment with EKS + Terraform
• Failure injection (pod kills, Redis outage, latency spikes) with written postmortems
• Then: secrets management (Vault), GitOps (ArgoCD), mTLS

• TFX-Addons (Outreachy 2021): built a custom schema component for TensorFlow Extended ML pipelines.
• Shell-C: a Unix shell from scratch in C, with piping, I/O redirection, process control, and signals.
• Contract backend work (2023–24): designed, deployed, and maintained a production MERN app for an automotive SaaS client, including billing automation and RBAC.
• Research: Maximal Secret Reconstruction, Teleportation and Bell's Inequality, EPJ D 79, 11 (2025).

Infra: Kubernetes · Helm · Docker · GitHub Actions · Prometheus · Grafana · Linux · DNS/TCP-IP Code: Python · Go · Bash · SQL · C Data: PostgreSQL · Redis · MongoDB

Remote platform / SRE / DevOps roles (full-time or contract), and freelance work on Kubernetes setup, CI/CD, and observability. Best way to reach me: email or LinkedIn.

_{cozy corners, hot tea, cold servers ☕}