The same project that took me four days, now built in an afternoon

Project Link : https://invoice-generator-teal-zeta.vercel.app/
Github Link : https://github.com/pratik23rj/invoice-generator

Can a designer actually build the thing?

I tried building an invoice generator about five years ago. It took me five days. The hardest part wasn't the logic, it was finding a library that would render the PDF exactly the way I wanted, and most of them got close but never right. I spent more time fighting render quirks than building anything useful. So when I picked the idea up again, the real question was whether the gap between designing something and shipping it had actually closed. Could I build a real product now, not just mock one?

The parts most invoice tools get wrong

Indian GST invoices have rules that look simple until you build them. The invoice has to auto-switch between CGST plus SGST and IGST depending on whether the buyer is in the same state. Numbers need Indian formatting with lakhs and crores, not the international comma style. The total has to be spelled out in words. And the preview has to update live as you type, so what you see is what gets saved. With Claude as a coding co-pilot, the work that used to be a five-day slog became an afternoon. The PDF render that broke me five years ago just worked.

Free, private, and shipped

The stack is Next.js and TypeScript, deployed on Vercel. There's no login, no backend, no subscription. Everything runs client-side, so the data never leaves the browser, which matters when people are typing in client names and tax numbers. It's free because the freelancers and small businesses who need it shouldn't have to pay a monthly fee to send an invoice. The point I keep coming back to is that the distance between an idea and a working product is shorter than it used to be, and I'd rather build the thing than wait for someone else to.