I build systems that hold under pressure. My work lives at the intersection of performance engineering and security architecture; where every design decision is deliberate and every abstraction earns its place.
From Go microservices handling concurrent event streams to React interfaces that surface complex data cleanly, I operate across the full stack with the same standard: robust, observable, and built to scale without surprise.
I'm an active open source contributor ; working across production-grade infrastructure projects spanning Kubernetes operators, distributed systems, security tooling, and scientific computing platforms. I contribute where the code actually runs at scale.
Security is not a feature I add after the fact. It's a constraint I design around from the first line ; production-grade authentication, cryptographic integrity, and failure-tolerant pipelines are the baseline, not the finish line.
Go-powered pipeline ingesting browser events through a stateful rule engine. Detects hesitation, navigation loops, and checkout abandonment in real-time. Live Astro/React dashboard with 2-second polling.
Three-service architecture ; React 18 SPA, Node.js API, Go authentication microservice. Containerized with Docker Compose. Production-grade security with RS256 JWTs, Argon2id hashing, Razorpay payments, and sliding-window rate limiting.
Production-grade atomic structure platform shipped as a single HTML file. Build atoms particle-by-particle with Three.js 3D visualization and real-time quantum mechanical simulations; progressively unlock all 118 elements through experimentation. Includes a Reaction Laboratory, ChemistDex compound encyclopedia (200+ entries with rarity tiers), and 50+ achievements with a 3D virtual trophy case.
Next.js 14 single-page experience with 11 interactive Canvas visualizations ; Fourier transforms, quantum wave packets, Lorenz attractors, Mandelbrot sets. Scientifically rigorous with an accuracy badge system.
Active contributor to production-grade open source infrastructure ; distributed systems, Kubernetes operators, security tooling, scientific platforms, and high-performance data pipelines. I go in where the code runs at scale.
I use AI to accelerate codebase onboarding ; rapidly mapping architecture, tracing call paths through unfamiliar systems, and constructing controlled environments designed to surface bugs before they reach production. The target is failure modes that cause real data loss, service disruption, or audience impact at scale: the kind of bugs that are invisible until the system is under load, and catastrophic once they surface.
Tap card to explore · Double-tap to zoom · Swipe to navigate
58 PRs across 21 distinct production systems means rapid, deep codebase onboarding across Go, Java, Rust, Python, and TypeScript ; every single time, cold, no handholding.
Cross-ecosystem fluency at this depth is rare. The same precision applies in Go schedulers, Java Thrift layers, Rust resolvers, and Python data pipelines. That's a way of thinking, not a list of languages.
These bugs don't surface in CI. They appear under production load with real data and real audiences. Finding them requires reading distributed systems code at the level it was designed ; not the level it was documented.
Concurrency bugs are non-deterministic, often silent, and only surface under specific timing conditions. Finding 8 across 5 different production codebases is a pattern of thinking ; not luck.
When cilium backports your fix to two stable releases, the bug was real, the fix was correct, and the reviewers trusted the code enough to ship it to production clusters running at planetary scale.
Whether it's a system that needs hardening, an architecture that needs rethinking, or a problem that needs someone who won't take shortcuts ; I'm listening.