A Non-Coder Amazon Seller Built His Own Tool in 48 Hours with Cursor, Then Traded Equity for Access to the Audience of a Coaching Program He’d Already Bought Into — Launch Fast / Hasaam Bhatti

Background & product

Launch Fast is an AI research/launch tool for Amazon FBA sellers. Feed it a keyword or ASINs and it grades a market with a proprietary A10–F1 score (A10 = best, F1 = worst), then surfaces historical price/BSR/revenue/review/ad-CPC trends, keyword comparisons, Alibaba supplier scoring, and financial modeling (COGS, freight, FBA fees, break-even) end to end. Beyond a Chrome extension and dashboard, it ships a remote MCP bridge so a seller’s own AI agents (Claude and others) can query Launch Fast’s tools and their Amazon data directly — an ‘AI-native’ design.

It was built by Hasaam Bhatti. Toronto-based, he’s been doing business online for about a decade (blockchain dev → Amazon FBA → AI/agents). He has no CS degree and had never written a line of code before Launch Fast. But he’s a domain expert in the day job: 7+ years selling on Amazon, 20+ products launched across Home & Kitchen, Baby, and Sports, running his own brand portfolio (‘HB Goodies’).

The trigger was the product research he’d done by hand — 20–30 hours per idea. Using the AI coding tool Cursor, he assembled a production-grade backend (TypeScript + Zod validation, retries, multi-layer caching) in 48 hours. His own GitHub repo describes it as turning ‘8-hour research into a 30-second AI conversation’ — that repo is actually the MCP server Launch Fast itself exposes to Claude and other AI agents, a separate thing from what he personally coded with.

And the real win wasn’t marketing — it was distribution. By his own account, two years before building Launch Fast, Hasaam had bought a coaching program called LegacyX FBA (six live calls a week plus 27 hours of video lessons) that already had thousands of active Amazon sellers. After building the tool in 48 hours, he pitched the LegacyX community: if they liked it, they’d partner — and he traded equity for formal access to that audience. He wasn’t building trust with strangers from zero; he delivered a new tool into the audience of a business he already had a stake in. That took him to $10K MRR in 30 days and $30K MRR within months. The official site says ‘990+ Amazon sellers’ use it.

The repeatable playbook

1. 1Hunt for ‘the boring manual task you’ve done for 7 years,’ not ‘the exciting thing you want to build’ — domain knowledge is the moat
2. 2If you already have a stake in some business or community, trade equity for formal access to its audience for your new tool — skip the slowest step of building trust with strangers from zero
3. 3A non-coder can ship a production tool in 48 hours with an AI coding tool (e.g. Cursor) — let AI do implementation; you own the spec via domain knowledge
4. 4Encode market judgment into a proprietary score (e.g., an A10–F1 grade) — turn your instinct into a formula so it isn’t a borrowed metric
5. 5Expose your features over MCP and bet on the new surface where ‘the user is an AI agent’
6. 6Sell a boring-but-painful B2B niche at $50–199/mo — high price makes revenue add up with few users

Deep dive

【Deep dive】The essence of Launch Fast isn’t the headline ‘a non-coder built it in 48 hours.’ It’s that three things line up: (1) who should build it (= be your own customer), (2) what made building possible (= an AI coding tool), and (3) how you distribute it (= trade equity for formal access to an audience you already had a stake in). Taken in turn:

■ (1) The strongest starting point is a chore you’ve done for 7 years. Hasaam wasn’t hunting for an idea. He turned an *existing* pain — the 20–30 hours of manual research he ran on every product idea across 7+ years of Amazon selling — straight into a tool. This works because domain knowledge itself is the moat. The A10–F1 grade that scores a market isn’t a borrowed metric; it encodes the instinct for ‘which numbers matter and how’ that he built launching 20+ products. A developer who’d never sold on Amazon couldn’t make that scoring credible. The lesson for indie builders is blunt: hunting for the ‘boring manual task you already do daily’ beats chasing the ‘exciting thing you want to build.’

■ (2) ‘I can’t code’ is no longer a barrier to entry. He has no CS degree and had never written a line before Launch Fast. Yet with the AI coding tool Cursor, he assembled a production-grade backend (TypeScript + Zod validation, exponential-backoff retries, multi-layer caching) in 48 hours. Crucially, AI filled in the *implementation*, not the *judgment of what to build*. The spec — which data, scored how, shown in what order — came from his domain knowledge; AI just made it fast. The structural shift is that domain experts can move from ‘the one who commissions software’ to ‘the one who ships it.’ Worth separating: Launch Fast the product does expose an MCP server to Claude and other agents, but that’s a different thing from what he personally coded with.

■ (3) Buy distribution — trade equity for access to an audience you already had a stake in. Most cases in this database built their audience *themselves* — TikTok virality, SEO, a personal following. Launch Fast is neither. By his own account, two years before building Launch Fast, Hasaam had bought a coaching program called LegacyX FBA that already had thousands of active Amazon sellers. So he wasn’t an outsider riding a stranger’s audience — after building the tool, he secured formal access to the audience of a business he already had a stake in, in exchange for equity. That’s why $10K MRR in 30 days was even possible: the slowest step, building trust from zero, was replaced by trust rooted in his own ownership stake. Translated for indie devs: if you already have a stake in some business or community, ask whether you can formally deliver a new tool into its audience via equity or revenue share — before trying to win over strangers.

■ (4) MCP-native design — betting that ‘the user is an agent.’ Launch Fast doesn’t stop at a dashboard; it exposes a remote MCP bridge so a seller’s own agents (Claude and others) can hit Launch Fast’s tools and their Amazon data directly. He bet early not just on ‘a UI where a human clicks’ but on ‘an API that an AI agent calls’ — a new product surface. As agents become the practical ‘users’ of software, whether your tool is exposed over MCP could become a real edge. It’s the same instinct as Postiz being first to ‘AI agents as a new kind of buyer,’ executed in a vertical B2B tool.

■ (5) Pricing and unit economics — the quiet strength of ‘boring B2B.’ Pricing is $50–199/mo (the top Scale plan at $199). That’s an order of magnitude above the $5–10/mo of viral consumer apps, so revenue adds up with fewer users. The official site cites ‘990+ Amazon sellers.’ Even if only a few hundred pay, the high price makes $30K MRR internally consistent. Sellers reason ‘$199/mo is cheap if it erases 20–30 hours of work and a bad-inventory bet,’ so willingness to pay is high to begin with. Unflashy, but a boring B2B niche with clear, specific pain is a ‘quietly strong’ market for a domain expert.

■ (6) Handling the numbers and the story (transparency). $30K MRR, $10K in 30 days, and the LegacyX-acquisition/equity-swap story are all self-reported in the founder’s Indie Hackers post, with no independent second source at this point (WebFetch returned 403 on the source URL, so this rests on WebSearch result summaries rather than a direct read of the full post). They don’t contradict the official ‘990+ Amazon sellers’ × official pricing ($50–199/mo), and the product, extension, GitHub, and founder site are all confirmed across multiple sources. This article adopts the self-reported figures conservatively while stating plainly that they are single-origin.