I'm a third-year engineering student at IITRAM, Ahmedabad.
In the last year I shipped 7 production AI apps and 7 open source libraries — all while attending college full-time. No team. No funding. No co-founder.
Here's exactly how.
Build silently. Ship publicly.
Most people announce what they're building, get dopamine from the likes, lose motivation when it gets hard, and quietly drop it.
I stopped announcing. I started shipping.
My apps are all private repos. No previews, no "coming soon" posts. I build until they're live, then I tell people. The libraries are the opposite — public MIT, on npm and PyPI, maintained in the open. That's the proof of work. The apps are the business.
Why libraries, not just apps
Apps are IP. Libraries are reputation.
Every time I extracted reusable logic from an app into a standalone library, two things happened — the app got cleaner, and I got a public artifact that developers could find and star.
My libraries cover things I genuinely needed and couldn't find done well: LLM routing by prompt complexity, Razorpay paywalls for React, offline-first sync engines, fraud detection you can pip install.
None of them are moonshots. All of them solve one specific problem, have zero runtime dependencies, and fit in a README.
The stack I stopped reconsidering
Decision fatigue kills momentum. I picked a stack and stuck with it.
React + Vite + Tailwind for the frontend. Supabase for the database. Groq for AI inference. Razorpay for payments. Vercel for deploys.
Every new app starts from this. I'm not re-evaluating frameworks every project. I'm building.
How I decide what to build
One filter: would I pay for this myself?
Not "would someone pay for this" — would I, personally, right now, pull out my card.
If yes, I build it. If no, I skip it no matter how clever the idea sounds. This filter has killed a lot of bad ideas fast and saved me hundreds of hours.
The mistake most student builders make
They spend 80% of their time building and 20% on getting users.
It should be 60/40 at minimum.
Building is the comfortable part. Getting strangers who don't know you to care about what you made — that's the hard part. Most people avoid it by adding more features instead.
Your product doesn't need one more feature. It needs ten more users.
What I've shipped
Libraries (all MIT, all on GitHub at github.com/iamadhitya1):
- llm-router — routes prompts to the right LLM by complexity and cost
- react-premium-gate — Razorpay subscription paywall for React
- groq-chain — Python LLM chaining with a .step() builder API
- react-offline-first — offline-first React template with IndexedDB sync
- fraud-shield — pip-installable fraud detection with Random Forest
- react-macro-rings — animated SVG nutrition progress rings
Apps: 7 live products under Rewrite Labs — all private, all monetized.
The honest part
Most days are just writing code, fixing bugs, and shipping things nobody noticed yet.
The compounding happens quietly. One library gets starred. One app gets a paying user. One article gets shared in a Slack channel you've never heard of. You don't see it building — then one day you look back and there are 14 live products with your name on them.
That's the whole strategy. Pick a problem. Pick a stack. Ship before you're ready.
Find everything I have built at https://github.com/iamadhitya1
Top comments (1)
Some comments may only be visible to logged-in visitors. Sign in to view all comments.