Two Boys From Town: From a Jewellery Shop to the Internet

This is the long version. The short version: 2 brothers, 22 years old, 4 failed businesses, 1 working one, 6 cohorts trained, 250+ builders, 700+ free AI skills, 100+ systems, 200+ resources. Built from the second floor of a gold shop in a town nobody's heard of.

The town.

Vaniyambadi. Tirupattur district, Tamil Nadu. Population ~100,000. Famous for: a leather industry that died in the 2000s, a 4-road junction, a 200-year-old dargah, and Maruthi Jewellers (where my family has been for 3 generations). Not famous for: AI, internet businesses, or anything we're doing.

This is the story of how we ended up doing AI from here anyway.

2018-2020: the failed businesses.

2018, 18 years old: a resume builder using GPT-3. Made ₹4,000 in 3 months. Quit when OpenAI changed the API. 2019, 19: a copywriting tool for small businesses. Made ₹0. Quit when I realised I was building a tool, not a business. 2020, 20: a "consulting" business where I'd help SMBs "adopt AI" with no actual experience. Made ₹8,000. Quit when I realised I was the customer, not the consultant. 2020 Q4, 20: a newsletter on AI in Indian SMBs. Got 47 subscribers, all friends. Quit when I realised I had nothing to say yet.

Four failures. Zero revenue. Most of my 20s' savings spent on Jio data and Claude subscriptions. But: I learned prompt engineering, API design, SMB outreach, content production, and the hard lesson that "AI tool" is not a business.

2021: 12th standard, St. Paul's.

OK, real story. Nav (me) and Sujal (brother) did our 12th standard at St. Paul's Matric Hr Sec School, Vaniyambadi, 2020-2021 batch. We were not the top of the class. We were the kids who did AI projects in the back of the notebook. The maths teacher let us use the computer lab after school. That's where the 2BFT name came from — the 2 of us, building things, while everyone else was studying for JEE.

2022-2024: the working one.

2022: we ran the first cohort of the 2BFT Academy. 8 people. 4 weeks. ₹2,000 each. We taught prompt engineering, API design, and the basics of building with LLMs. The cohort finished. 6 of 8 shipped a working project. ₹16,000 in revenue. Not life-changing, but it worked.

2023: 4 more cohorts. 2024: 3 more. Total 6 cohorts, 250+ builders, 30+ shipped products. By 2025, our 8th and final cohort was the biggest — 60 builders across 3 cities (Vaniyambadi, Chennai, Coimbatore). We ran the Vibe Coding Hackathon at Zoho Corporation in June 2026, free, 12 teams, all shipped. We stopped running cohorts and started the Academy at 100x scale.

Why we stopped cohorts.

6 cohorts taught us that the cohort model is the wrong shape for India. It works for 30-50 builders per batch. It doesn't scale to 5,000. So we built the Academy: 4 phases, 18 lessons, 17 hours, free Phase 0, paid Phases 1-4. Same content, 100x scale. The 250+ alumni are the proof it works. The 700+ free skills are the gateway.

Where we are now.

July 2026: 2BFT is the most-visible AI education brand in Tamil Nadu. We've worked with Zoho Community, EDII Tamil Nadu, and St. Paul's school. The 2BFT Academy is free for Phase 0. The 200+ resources on /learn are free. The 100+ systems on /agentic-systems are free. The 3 services (Custom Workshop, UGC, Agents Setup) are paid. Stashed ships. Maruthi Jewellers is still downstairs.

What we believe: the next 100M AI users in India will come from tier-3 towns, not metros. They'll learn from free, plain-English resources, not ₹50K bootcamps. They'll build on free tools, not paid SaaS. They'll serve local customers, not global markets. That's the play. That's 2BFT.

Welcome to the floor above the gold shop. Sit down. There's work to do.