How I built 3 products solo as a CA student using AI — no coding background

A CA Final cleared student from Mumbai, Aakash, has built and shipped three live software products in six months without any coding background, using AI tools like Claude and Cursor to write the code. The products include a full-stack SaaS bookmark manager, a PDF management tool, and automated F&O trading software, all developed solo with zero funding. Aakash attributes the success to focusing on product decisions and deep domain knowledge from his CA background, while AI handled the actual code.

I'm a CA Final cleared student from Mumbai. I don't have a computer science degree. I never took a coding bootcamp. I couldn't write a for-loop from scratch six months ago. And yet — I've shipped 3 real products, solo, with zero funding. This is exactly how I did it. It was late at night. My laptop crashed. When it came back on — 3 years of saved research was gone. Bookmarks, references, articles, everything. Just wiped. I went looking for a tool to make sure it never happened again. I couldn't find one that worked the way my brain does. So I decided to build one. There was just one problem. I didn't know how to code. Vibe coding is building software using AI tools like Claude, ChatGPT, and Cursor to write the actual code — while YOU focus on the product decisions. You describe what you want. The AI builds it. You test it, break it, fix it, ship it. Is it "real coding"? Honestly — I don't care about that debate. What I care about is this: 3 products. All live. All real users. Built by someone who couldn't write a for-loop 6 months ago. That's real enough for me. The problem: Every bookmark tool stores links. None help you find them again. What I built: A full-stack SaaS with: Stack: Next.js · Supabase · Tailwind CSS · Vercel Time to build: 6 weeks Lines of code I wrote myself: Honestly? Almost none. Claude wrote most of it. Is it live? Yes. Launched Product Hunt May 2026. The part I contributed: → The product idea → Every design decision → Every UX choice → Testing every single feature → Debugging when AI got it wrong → Shipping it The problem: PDF tools are either too expensive or too complicated. What I built: A clean, fast PDF management tool for people who work with documents daily. As a CA student — I live in PDFs. Tax documents, audit reports, financial statements. I built this because I needed it. Time to build: A few weeks Biggest lesson: Sometimes the best products are built for yourself first. This one is my favourite. The problem: F&O trading software is either too expensive for individual traders or too simple for serious ones. What I built: Automated F&O trading software for both institutional and individual traders — with real-time market execution and risk management. Why I could build this: My CA background meant I actually understood the finance side deeply. The AI handled the code. I handled the logic. This is where being a CA + a builder became a genuine superpower. Lesson 1: The idea is the hard part. AI can write code. AI cannot tell you what to build or why it matters. That's 100% you. Lesson 2: You need to understand the problem deeply. Lesson 3: Testing is your real job. AI writes the code. You break it. Then you describe how it broke. Then AI fixes it. That loop IS the job. Lesson 4: Ship ugly. Fix fast. My first version of Smart Bookmark Manager was embarrassing. I shipped it anyway. Real users found bugs I never would have. Lesson 5: Your unique background IS your moat. AI Coding: Claude Anthropic Cursor IDE Frontend: Next.js + Tailwind CSS Backend: Supabase Hosting: Vercel Design: Canva + v0.dev Learning: YouTube + Stack Overflow when AI got confused Total monthly cost: Under ₹2,000 Smart Bookmark Manager just launched on Product Hunt. I'm working on: → Browser extension → AI categorisation Pro → Web archiving Ignis Quant is being refined for more traders. PDFly is getting a UI overhaul. And I'm probably going to build a 4th product. Because I can't stop. You don't need a CS degree. You don't need to know how to code. You need a problem worth solving. You need patience to test and iterate. You need the courage to ship something imperfect. The tools to build anything exist today. They're free or nearly free. The only thing stopping you is starting. So start. I'm Aakash — CA Final cleared, solo builder, Mumbai. Currently building Smart Bookmark Manager. Would love your feedback and thoughts in the comments below 👇 Try it free: smartbookmarkpdfmanager.vercel.app