NASA Space Apps Winner
Regional — built a fully working AI prototype in 48 hours
Full Stack Engineer with deep experience across React, Next.js, FastAPI, AWS and Docker. I ship scalable APIs, AI integrations and cloud-native deployments — and obsess over performance and DX.
Currently an Associate Full Stack Engineer at Snapshare AI. I love shipping production systems where the bottleneck moves between frontend perf, API latency, and infra design — and getting every one of them right.
I'm a Full Stack Engineer who builds production-grade systems end-to-end — React/Next.js on the front, FastAPI & Django on the back, Docker + AWS in between. I take features from a Linear ticket to a load-balanced container in EC2 with CI/CD on every push.
At Snapshare AI, I cut React bundle size by ~40% and slashed average API response time by 60% through Redis caching, DB indexing and pagination. I won the regional NASA Space Apps Hackathon by shipping an exoplanet-classifying AI in 48 hours.
I care about the seams: where frontend perf meets API design, where dev ergonomics meets deploy speed. The best systems feel quiet — that's the bar.
Drag to spin · Click a tag to see related projects. Color-coded by domain — indigo for languages, pink for frameworks, mint for cloud, amber for databases.
From a Django shop in Kochi to a remote-first AI startup in Bangalore — every role tightened the loop between idea and production.
A computing track from BCA through MCA — each step deepened the systems and engineering side of how I work today.
Kannur University · Department of IT
Kannur University
Science
IMNSGHSS Mayyil
Each card is something I designed, built and deployed — not a tutorial, not a fork. Tilt them with your cursor.
AI-Powered Interview Training
Complete interview-training platform with facial emotion analysis (FER2013), Gemini-powered feedback and answer evaluation, and speech-to-text / text-to-speech for interactive interviews. Dockerised for consistent deployment.
NASA Space Apps Hackathon Winner
AI system built in 48 hours to classify exoplanets from NASA datasets. ML model engineered on 12 planetary features with 2D orbital visualisation and habitability analysis. Working prototype shipped under extreme time constraints.
Full-Stack AI Chat App
AI-powered chat application with real-time messaging, Gemini LLM with context retention, JWT auth and session validation, rate limiting, Redis caching and chat-history management.
AI Pothole Detection
YOLO-based pothole detection with A* pathfinding for safe-route calculation. Flask REST APIs for inference, React real-time visualisation, fully containerised inference service.
I'm always up for collaborations on AI-powered products, cloud-native systems and ambitious side-quests. Drop a line — I answer fast.