He Learned AI in One Weekend, Then Built a $15K-a-Month Chatbot Solo — No Team, No Funding

Process

Bhanu Teja P — a solo developer from India — built SiteGPT, an AI customer-support chatbot you train on your own website. With no team and no funding, he reached $10,000 in monthly recurring revenue in 30 days and about $15K MRR within months — from a startup cost of essentially $0.

Stage 1 — Quit the job, move home, and ship into the void for years

Bhanu Teja P is an IIT engineer — and on paper, he had the safe path. Instead, after just 8 months as a software engineer at Swiggy (one of India's biggest startups), he quit, moved back into his parents' house to drive his expenses close to zero, and went full-time on building his own products. It wasn't a glamorous decision; it was a calculated one. Living rent-free bought him the single most valuable thing an indie maker can have: time to fail. And fail he did — for more than three years he shipped product after product, most of them going nowhere. But each flop was quietly teaching him the three skills that matter: how to ship something real, how to find an audience, and how to actually sell.

Stage 2 — The first real win: Feather, $6K MRR, sold for $250,000

After years of small bets, one finally landed: Feather, a tool that turns a Notion page into a real blog. He grew it patiently to $6K in monthly recurring revenue over about 15 months, then sold it for $250,000 — a 3.5x annual-revenue multiple. That exit did two things. It proved the years of "failure" had actually been an apprenticeship — he now knew how to take a product from zero to a paying audience to a clean sale. And it bought him the freedom and the confidence to make a bigger, faster bet when the moment came.

Stage 3 — A weekend to learn AI, two weeks to launch SiteGPT

In early 2023, his Twitter feed filled up with AI tools overnight. Most people just watched the hype roll by. Bhanu instead set aside a single weekend to learn the new tech the only way that sticks — by building something useful with it. The result was SiteGPT: an AI chatbot you train on your own website's content, so it can answer your customers' questions automatically. He aimed a brand-new capability (AI) at a boring, universal need (customer support). From idea to launch took just two weeks. He posted it to Hacker News's "Show HN," where it hit the front page; two days later he launched on Product Hunt and took #1 Product of the Day. The early traction was immediate.

Stage 4 — $10K MRR in 30 days, ~$15K within months — built on free tools, not ads

SiteGPT hit $10,000 in monthly recurring revenue in its first 30 days and climbed to about $15K MRR within months — with zero marketing budget. His secret weapon was "engineering as marketing": instead of buying ads, he built 50+ small free tools around his product's topic, and those tools pulled in 90% of his Google traffic — for free, and for years. He shared everything he built on Twitter as he went, the same build-in-public muscle he'd trained on Feather. All of it solo, no funding. He later sold SiteGPT in a micro-acquisition (paid out in two tranches over six months) and now runs other projects, clearing $20K+ MRR combined.

"My best marketing is building things and sharing what I build. Build 50 free tools, and the traffic takes care of itself." — Bhanu Teja P (paraphrased from public interviews)

Source: Indie Hackers · Mixergy · Superframeworks · Bhanu Teja's public posts (@pbteja1998)

Thinking

Insight 1: Ride the wave the week it forms — a weekend of learning can be enough to ship

When AI tools flooded his feed, he didn't just watch — he spent a weekend building. The weeks a new wave is forming are a rare window: a little learning plus fast hands is enough to get on board before the crowd.

Insight 2: "Engineering as marketing" — build free tools that do your marketing for you

He built 50+ small free tools that brought in 90% of his Google traffic, at zero ad spend. Turn your marketing into a product: free tools keep pulling in people long after an ad would have stopped.

Insight 3: Years of "failures" are the tuition for the one that works

Quitting his job, moving home, shipping flops for years — that's what taught him to ship, find an audience, and sell. The "overnight" win sits on years of practice nobody applauded.

Insight 4: Productize a brand-new capability for a boring, universal need

He took a brand-new capability (AI) and aimed it at a boring, universal need (answering customers' questions). New tech times an old, unglamorous need equals a real business.

Insight 5: Build and sell — small exits compound into freedom

Feather sold for $250K; SiteGPT went in a micro-acquisition. A string of "build it, then sell it" wins compounds into more freedom and bigger bets each time.

Action

Step 1: When a new capability floods your feed, spend a weekend learning by building

Don't just bookmark the hype. Pick one new capability and build something tiny but useful with it over a weekend — shipping teaches faster than watching.

Step 2: Aim the new tech at a boring, universal need

Don't chase clever — chase needed. Find an unglamorous problem lots of businesses have (support, FAQs, onboarding) and point the new capability straight at it.

Step 3: Go from idea to launch in weeks, not months

Set a two-week clock. Launch on Show HN and Product Hunt — they're free distribution that rewards something real and rough over something perfect and late.

Step 4: Build free tools as your marketing engine

Instead of buying ads, build small free tools around your product's topic. They rank on Google and pull in the right people for years — marketing that compounds.

Step 5: Build in public, and don't be afraid to sell

Share what you build on one platform, and when an offer is good, take the exit. Each "build it, sell it" round buys you freedom and a bigger next bet.

Not for you if: you won't learn to build software (this path needs code — though AI is lowering that bar fast); you want passive income (free tools, launches, and support are real work); or you expect the first thing to hit (his win sat on years of flops).