Somehow, I get
paid to make
machines think.
I build backend & AI systems that handle real-world chaos without panicking — most of the time. Agent workflows, distributed tracing, and the occasional 3 a.m. production crime scene.
Things I built that actually shipped.
A short collection of systems that made it to production and survived contact with real users.
ShellBreaker
Linux terminal training ground
A platform where breaking production is the point. Practice Linux commands, debug chaos, and master the terminal in isolated environments — because textbook Linux is clean, but production is a crime scene.
Things I'm genuinely good at.
A range of technical skills — mostly sharpened by solving problems I created for myself.
Languages
Frameworks
Databases
Architecture
Real-Time & Async
Cloud & DevOps
Places that trusted me with their code.
A timeline of teams that let me break things until I learned how to fix them.
Building at an early-stage YC S26 startup, shipping product and infrastructure end to end. Working close to the metal on backend services and AI features, moving fast, owning problems, and turning half-formed ideas into things that run in production.
Designing and scaling backend infrastructure for AI agent workflows handling multi-step, long-horizon task execution in production. Implemented structured observability pipelines — distributed tracing, structured logging, and alerting — to surface failure modes and latency regressions across agent runs. Built fault-tolerance mechanisms including circuit breakers and graceful degradation for unreliable tool and API dependencies. Collaborating directly with founders to define system architecture, infrastructure decisions, and engineering roadmap.
Building a RAG system for finance teams to query CRM, Excel, and Salesforce data using LLMs and FAISS. Designed backend, vector search, and AI pipelines with a focus on modularity and future scaling. Contributing at a fast-paced fintech startup led by ex-Optiver, Tower Research, and Goldman Sachs founders.
At Scaler, I built an AI support chatbot that reduced human workload (and excuses) by 33%, thanks to vector databases and some backend automation magic. Created a universal AI suite to juggle multiple LLMs without breaking things (most of the time), and used Langfuse to keep track of the chaos. Also built a notification system for WhatsApp and email – because apparently people like being notified when stuff breaks.
The ongoing academic struggle.
Still attending classes, submitting assignments, and pretending to understand complex algorithms.
Currently learning how to build real-world software — while battling deadlines, bugs, and the occasional existential crisis. They call it “practical computer science education.” I call it “professional problem creation (and sometimes solving).”
Pursuing a CS degree focused on theoretical foundations and practical applications — or as I describe it, learning things I’ll Google again before every exam.
Things that made me look cool online.
Milestones and recognition picked up along the way.
iQOO Hackathon 2026
iQOO × Scaler School of Business · Bengaluru · June 2026
Took the grand prize at a 30-hour AI build sprint — built an on-device 3D physics simulation engine that turns dry study material into interactive, explorable content. Prize pool cracked: ₹1,50,000.
Emerging Tech Hackathon
Metadome.ai · Bengaluru · 2024
Took 2nd place — solo — with a fully local, on-device image search engine that finds your photos without a single byte leaving the device. Built the whole thing myself. Prize: ₹50,000.
LinkedIn Presence
Built some tech projects and somehow tricked the media into thinking I’m a genius. Got featured in Times of India, Deccan Herald, Business Standard, and a few others — probably because they ran out of real news that week. Still counts though.
Mom's proud — I'm in the news.
Outlets that thought my projects were innovative enough to feature. Still waiting on the movie deal.
Stuff I built when I was ‘free’.
What happens when too much free time meets too much curiosity.
An innovative system that allows players to control racing games using hand gestures, providing an immersive and intuitive gaming experience.
A real-time sign language to English converter using Python and TensorFlow, enabling seamless online conversations for Deaf individuals.
Developed a system for detecting personal protective equipment (PPE) kits using computer vision techniques.
Created a logo detection system for identifying and tracking brand logos in images and videos.
Built a full-stack hotel room booking system with AI-powered chatbot assistance for customer inquiries and bookings.
Let's talk.
A project idea, a question, or just a hello — I reply within one working day.