Software engineer with an intentionally broad cross-section: production infrastructure at Oracle (cluster controllers, FIPS, DNS, scheduling), ML pipeline engineering in open source (TFX/TensorFlow), quantum computing research at IIIT Hyderabad's CQST, and currently building Ferrum : a production-grade webhook relay and delivery platform from scratch, as a structured deep-dive into DevOps, SRE, and platform engineering.
I operate across the full stack on purpose, from writing a Linux shell in C to deploying multi-region K8s clusters with zero-downtime SLOs. My goal is systems that hold at scale, not just systems that run in demos.
A self-initiated, 3-phase production DevOps build — designed to mirror real-world SaaS infrastructure challenges at scale.
Ferrum is not a tutorial project. Each phase targets a distinct engineering discipline, with real constraints and expert-level failure scenarios baked in.
Core DevOps: containers, CI/CD, Kubernetes, observability
| Architecture | API + Frontend + DB + Cache + Event-driven Queue |
| Tools | Docker · Kubernetes · Terraform · GitHub Actions · Helm |
| Scale targets | 50–100K users · zero-downtime deploys · horizontal + auto-scaling |
| Expert challenges | Simulate latency spikes + pod failures · debug with traces, not logs |
SRE + Security: zero-trust, secrets management, service mesh, incident response
| Architecture | Multi-region active-passive |
| Tools | Vault · Istio · OPA · ArgoCD · Prometheus |
| SLO targets | Region failover < 60s RTO · mTLS + policy enforcement |
| Expert challenges | Simulate DNS failure + cert expiry · run full incident + postmortem |
Internal dev platform + Kafka pipelines + FinOps
| Architecture | Multi-cluster + event-driven + self-service |
| Tools | Backstage · Kafka · Crossplane · OpenTelemetry |
| Scale targets | 1M events/min ingestion · multi-tenancy isolation |
| Expert challenges | Chaos engineering (network partitions) · enforce budgets on cost explosion |
INDUSTRY OPEN SOURCE RESEARCH SYSTEMS
──────── ─────────── ──────── ───────
Oracle SWE-1 TFX-Addons IIIT-H CQST Shell-C (C)
Cluster ctrl (tensorflow/tfx) Quantum Info Linux Internals
FIPS / DNS Outreachy Intern Algorithms DNS / Networking
Production Infra Custom TFX Component MS by Research K8s Scheduling
Distributed Sys Feature Engineering Quantum + CS Production Infra
Scheduling ML Pipelines Open Research Distributed Sys
ML STACK WEB STACK DEVOPS STACK SCIENCE
───────── ───────── ──────────── ───────
TensorFlow React Docker Quantum Info
TFX Pipelines Node.js Kubernetes Qiskit
Scikit-learn MongoDB Terraform Astronomy
Feature Eng. REST APIs GitHub Actions Data Analysis
Python/Jupyter Google OAuth Helm / ArgoCD Galaxy Clusters
ML Production MERN Full Stack Prometheus Actuarial Math
Grounded in actual work, not aspirations:
| Role | Evidence |
|---|---|
| Platform Engineer / SRE | Ferrum (multi-region, Vault, Istio, Prometheus, chaos engineering); Oracle (cluster controllers, scheduling, FIPS, DNS at production scale) |
| MLOps / AI Infrastructure Engineer | TFX-Addons contributor — ML pipelines at Google-scale; feature engineering, TensorFlow in production; Outreachy open-source track record |
| Backend / Distributed Systems Engineer | Oracle distributed systems (FIPS, DNS, production infra); Ferrum event-driven queue + 50–100K scale; K8s internals + scheduling |
| DevOps / Cloud Engineer | Ferrum Phases 1–3 (Docker → K8s → Terraform → IaC → FinOps → Chaos); CI/CD via GitHub Actions; Helm; ArgoCD |
| AI Software Engineer | Quantum + ML crossover (IIIT-H CQST, MS Research); open-source research credibility; full ML production pipeline experience |
Production-grade webhook relay and delivery platform. Built across 3 phases: Core SaaS DevOps → SRE/Security → Platform Engineering. Targets 50–100K users, sub-60s regional failover, 1M events/min. Built to learn by doing — the hard way.
Open-source ML pipeline contributions via Outreachy. TFX-Addons: 130+ ⭐. Built real TFX components — schema validation, feature engineering — used in production ML workflows.
ML applied to astronomical datasets — cross-matching, classification, data-driven discovery at cosmological scale.
🐚 Shell-C
Fully functional Linux shell in C, from scratch. I know what happens below the abstraction layer.
Dual-role auth (Applicant + Recruiter), Google OAuth, fuzzy search, resume management, full REST API backend.
- 🔩 Building Ferrum — self-directed, 3-phase production DevOps arc (K8s → Vault → Kafka → Chaos Engineering)
- 🏢 SWE @ Oracle — production distributed systems: cluster controllers, FIPS, DNS, scheduling
- 🌍 Outreachy Intern — contributed to Google/TFX's open-source ML infrastructure
- ⚛️ MS by Research — IIIT Hyderabad, Centre for Quantum Science and Technology
- 📦 Arctic Code Vault Contributor — code preserved for future generations
- 🦈 Pull Shark ×2 — consistent open-source contributor
- Ferrum Phase 1 — K8s + Terraform + GitHub Actions CI/CD pipeline; observability-first (traces > logs)
- Quantum-classical hybrid algorithms and NISQ-era applications
- Actuarial mathematics and risk modelling
Platform engineering, MLOps, distributed systems, quantum computing, or a good tea recommendation.
⚛️ quantum · 🔧 platform-eng · 🤖 mlops · 🌌 distributed-systems · ☕ caffeine