CS student at PES University building secure, user-first products and practical AI systems.
I build things to solve problems I genuinely face. When existing options feel fragmented, expensive, impersonal, or simply not good enough, I enjoy understanding the problem deeply and building the tool I wish existed.
My process starts with the customer journey and works all the way down to system architecture and implementation. I care about security, thoughtful user experience, and the overlooked features that make a product feel built with its users, not just shipped to them.
Software Engineering Intern — Mexo Technologies · 2026–Present
Building agentic workflows around manufacturing data to surface operational insights and support downtime reduction.
Software Engineering Intern — Cogniquest Technologies · 2025 · 2 months
Worked on document-intelligence workflows, including financial-table classification and structured data extraction.
Research Intern — RAPID, PES University · 2024 · 2 months
Built ML-driven systems, including a diet recommender and regression-based prediction work.
A free, frictionless resume and cover-letter builder for students.
I built it after having to pay hundreds of rupees for a simple resume edit because the tool I used made it difficult to change otherwise. Building a resume should be easy, practical, and never feel like a paywall trap.
React · Node.js · MongoDB · Google Cloud
A local-first AI assistant for everyday workflows.
I believe increasingly capable personal machines will make local AI a meaningful alternative to sending every interaction to a centralised data centre. CloneAI explores that future: a private assistant that can help across mail, calendar, voice, web search, documents, and image generation.
Python · Ollama · LLMs · Agents
A private home server for the information a family actually needs.
Family information should not be buried in drawers, known only by whoever manages it, or require a chain of calls to locate. FamilyBox began as a secure document home app and is growing into a local system for shared family records, health tracking, finance tracking, and more integrated seamlessly with a custom iOS app.
It is also my playground for experimenting with new technologies through features that solve real household problems.
iOS . App . Local-first · Encryption · Home server · Personal data
Interactive system-design practice built around real-world simulations.
As AI makes implementation faster, the differentiator between a good engineer and a great one will be the ability to reason about systems. Most system-design resources are theory-heavy, boring to work through, and difficult to visualise. SystemForces makes the learning process interactive, approachable, and gamified.
TypeScript · React · System design · Distributed systems
How I approach a problem
- Start with the real user workflow—not a framework or model.
- Find the friction, constraints, and moment that matters most.
- Design the experience first, then work downward through data, architecture, reliability, and implementation.
- Prefer simple, deterministic systems where correctness matters; use ML or LLMs where they add genuine judgment or unlock unstructured data.
- Keep refining until it is something I would be happy to use myself.
- Local-first and privacy-preserving AI
- AI agents for real-world operational workflows
- System design and distributed systems
- Applied ML, document intelligence, and causal reasoning
TypeScript · JavaScript · Python · React · Node.js · Flutter · MongoDB · SQL · Docker · GCP · Ollama
🧠 ~1.5B AI tokens spent turning ideas into products