Full-Stack Developer & AI/ML student who likes building things that work — from web apps and browser extensions to ML models and hardware projects.
I'm a B.Tech Computer Science student at Graphic Era Hill University, specializing in AI & Machine Learning. I like building stuff — whether it's a web app, an ML pipeline, or a Raspberry Pi robot.
I've freelanced as a full-stack dev, built a few projects I'm proud of, and competed in a couple of hackathons. Still learning a lot, but I enjoy the process.
Outside of code, I volunteer with NSS and occasionally tinker with hardware projects.
Full-Stack Developer & Automation Engineer
Codebase & Developer Intelligence Platform
Surfaces code ownership, technical-debt signals, and cited answers pulled straight from your repos and team chat. NestJS + BullMQ ingestion workers feed a pgvector-backed Postgres store behind a Next.js dashboard. Internally codenamed Aether.
AI Chess Tutor
Explains the consequence of every move, predicts what you were trying to do, and turns your own games into tomorrow's lesson. Runs Stockfish analysis and a BYO-key AI coach entirely client-side, against an in-browser mock backend built from real OpenAPI microservice contracts.
Browser Extension for Zomato Review Automation
A browser extension that automates collecting customer reviews from Zomato and generates intelligent, context-aware replies — streamlining restaurant reputation management.
Real-Time Messaging App
A real-time, one-on-one and group messaging app in the same product category as WhatsApp or Telegram — OTP + Google OAuth login, group roles, reactions, replies, and file sharing over a Socket.IO pipeline backed by MongoDB.
Autonomous Campus Tour Robot
Designed an autonomous robot on Raspberry Pi with sensor fusion achieving 95% obstacle avoidance. Implemented RAG pipeline using Pinecone & Gemini API for location-specific queries in under 2s.
Hand Gesture Recognition System
Trained a gesture recognition model with TensorFlow & MediaPipe achieving 90% accuracy at 24 FPS. Optimized inference by 35% through quantization for real-time Raspberry Pi deployment.
The layers that power everything I build
V-IPDS: A Vision-based Insect Predator Detection System for Beehive Monitoring Using YOLOv11n (2025)
2025Led a 4-member team to the finals, presenting a working prototype to industry judges
2025Advanced to finals in a national-level hackathon, competing against 500+ participants across India
2024Progressed from participant to coordinator, organizing community service drives and outreach programs
2023 — PresentI'm always open to discussing new projects, creative ideas, or opportunities to be part of your vision.