Ideas · Open to Co-Founding
Things I'd Build
Concepts I'm turning over. Some I'll take solo all the way to launch. Others I won't, because the right partner changes the product more than another commit would. The door's open if you're one of them.
Building with me
I'm open to co-founding.
Some of the ideas below I'd build as solo projects. Others I won't, because the fun part isn't the engineering, it's the users, the commercial motion, and the taste that only a partner brings. If you're already working on something and need a technical founder who can ship a full product end-to-end, or you have an idea that needs a builder, this is the door.
What I bring
- End-to-end build speed. Full-stack (Next.js, Prisma, Postgres, TypeScript), mobile (Swift, React Native), ML (Python, PyTorch, TF), real-time systems (WebSockets, MV3 extensions), Claude / Gemini / OpenAI glue.
- Commercial judgment from a real business. I run the full-stack platform for my family's 150-year jewelry business, real customers, real P&L, real inventory, real complaints.
- Shipped products, not prototypes. Watch Together is live on the Chrome Web Store. Gondilal is at gondilalsaraf.com. PCOD Tracker is in production. I finish things.
- The “know when to stop” muscle. I shelved Vaani after a market-fit check showed the unit economics didn't work. I'd rather kill an idea than force-ship it.
What makes a good partner
- You've talked to users. A concrete segment and a specific problem they've paid or would pay to solve. Not “Uber for X” energy.
- You cover the muscle I don't. Distribution, sales, design, a specific-domain wedge, the parts where “just ship it” isn't the bottleneck.
- You move. I ship full products in weeks, not quarters. A partner who can match the speed on their side is table-stakes.
- Not crypto / gambling / dropshipping / attention-scam. Life's too short.
On my desk
Fair Ludo: Provably Fair Dice Game
A consumer Ludo app where the dice are provably un-riggable.
Ludo at scale on every Indian platform is infamously rigged, operators tune dice to keep losing players hooked and new players lucky. I want to build the opposite: a consumer-grade Ludo app (web + iOS + Android) where every roll is server-authoritative under a commit-reveal scheme. The server publishes a hashed seed before each game and reveals it after, so any player can replay and mathematically verify every roll in the match. Core RNG + game logic open-sourced so the fairness claim is independently auditable. Planned stack: Next.js for web, React Native / Expo for mobile, shared TypeScript game engine, Node + WebSockets backend, Postgres for accounts and match history. Big-app ambition, real-money tournaments, matchmaking, replay sharing, social. Still pre-build; the technical design is the easy part, the consumer-growth motion is where a co-founder changes everything.
Co-founder note
Looking for a co-founder on this one, consumer-growth / marketing / Indian-gaming-market partner who can match the build pace.
Style It: AI Wardrobe Assistant
A digital wardrobe that actually helps you dress.
A software tool that maintains a digital wardrobe, each garment tagged with fabric, colour, care instructions, and photos of both the garment and the user wearing it. The goal is to solve the everyday 'I don't know what to wear' dilemma, reduce impulse purchases, and help people actually use the clothes they already own. Long-term hooks: weekly outfit planning that load-balances across fabric and colour to prolong garment life, a digital try-on loop, crowdsourced garment ideas with royalties for creators whose designs get produced, and affiliate-style recommendations. Spiritual successor to the earlier Style It experiment from high school, same problem, bigger frame.
CardRanker: Credit Card Value Calculator
Ranks credit cards by net annual value for a given spending profile.
SwiftUI iOS tool that ranks credit cards by net annual value given a spending profile across groceries, dining, travel, and other categories. Takes reward rates per card and annual fees, then surfaces the card that earns the most after fees rather than the highest sticker rate. Built as a Swift learning exercise. Shelved as a shipped app.
Also pitch me
Have an idea that isn't on this page? Even better. The ones listed are mine, the ones you bring are yours. If you have concrete users in mind and a problem worth building against, send me the sharpest version of your pitch and we'll take a day to talk it through.
Reach outThe bar to leave this page for Projects is a working prototype. Until then, everything here is still a question, not an answer.