A few things I’ve built — some to solve real problems around me, some simply for the love of building. Open any of them and try it for yourself.
The very first thing I built for people other than myself. I live in Raheja Vistas Elite, Nacharam, where the daily office commute was a shared headache — one that affected me too. So I built RVonWheelz to fix it: neighbours offer or book car and bike rides, chat with co-riders, and get realtime alerts for seat requests, approvals, and reminders. A small, non-profit solution to a very real, very local problem.
My most recent build — and the one I’m still actively tinkering with. It’s an AI companion for the 2026 World Cup: AI-written match stories, tournament simulations, and visual analytics, with a live Ballon d’Or prediction running right now. Built offline-first as a progressive web app, so it installs straight onto your phone like a native app.
This is where I pushed the AI the furthest. Invest.IQ leans on heavy analytics and an AI reasoning layer to turn a tiered universe of Indian equities into a single, opinionated, backtest-first shortlist across short, medium and long horizons. Quality scores, technicals, and news influence are all weighed — and explained — so you can see exactly why each pick ranks where it does. It’s a personal project and my playground for real-world applied AI. Educational only — not financial advice.
Built for the people who worry when you travel. Koode (Malayalam: “together / with you”) lets family and friends know you’re okay without you having to call or text. It doesn’t track you — it watches the journey, rating its health as Normal, Attention or Concern, and gives loved ones a calm reassurance channel with live map, ETA and timeline. Private by design: invitation-only, and every trip self-destructs 30 minutes after you arrive. Fully open source and zero cost — OpenStreetMap for maps and Supabase for live sharing, no Google, no billing anywhere.
A demo I built to show what a real “knowledge fabric” looks like — the kind of explainable, citation-grounded AI a lot of teams are chasing right now. Every answer is traced back through document → page → section → paragraph, and a live knowledge graph shows exactly how it was derived. Fully static: no backend, no vector database, no API keys — just open it and watch the reasoning unfold.
My most personal project — a hobby I started simply to make life easier for my mom and my wife in the kitchen. It looks simple, but it’s quietly clever: a 100% offline South-Indian meal planner that suggests today’s meals from what’s actually in your kitchen, remembers festival menus, and tracks grocery and household reminders. No accounts, no cloud, no tracking.
One of my favorite sports side projects. I built IPLBuzz to keep the entire IPL 2026 season in one place — from points tables and squads to batting and bowling leaderboards. What started as a simple season tracker grew into a complete cricket analytics dashboard and a fun way to explore tournament data.