Sold AI Writing Tool for $1M, Got Bored, Built AI Headshot Site in 30 Hours — $100K in Two Weeks
Dutch developer Danny Postma built a design resource site at 21, sold AI writing tool Headlime for $1M, took time off until boredom struck, then built HeadshotPro in 30 hours when Stable Diffusion launched — $100K in two weeks, $300K MRR within a year.
Process
At 21, he built the thing he couldn't find
Danny Postma didn't start as a programmer. In his early twenties, he was doing web marketing and conversion rate optimization in the Netherlands — testing landing pages, reading analytics, a practical digital marketer solving measurable problems for clients.
In 2015, he was 21, and a problem kept coming up: when he needed to build a landing page, there was nowhere good to look for examples. Inspiration was scattered across the internet without a single organized reference library.
So he built one himself. He called it Landingfolio — a curated collection of landing page designs. He didn't know how to code then, so he built it on WordPress.
Landingfolio grew slowly, but it found users. More importantly, it sharpened a principle Danny would follow for the next decade: find something you genuinely need that nobody has done well, and make it.
Chapter two: AI writing tool, $1M exit
In 2020, GPT-3 opened. Danny saw an opportunity. He built Headlime — a tool using AI to help marketers write ad copy, email drafts, and landing page content faster. This time he learned enough code to build it himself.
Headlime reached $20,000/month in revenue before being acquired in 2021 for $1,000,000.
He took time off. It sounded ideal.
But leaving a company you've built turns out to be a strange experience. All the decisions, problems, and user emails that had filled his days — gone overnight. He spent a while not knowing what to do next.
In an interview, he described that period simply: "I was a bit bored."
September 2022: Stable Diffusion was released to the public
Stability AI opened their image generation model to anyone who wanted to use it. This was a genuine inflection point — a new technical capability that anyone could build on top of.
Danny was living in Bali when the news hit. His first instinct was to build a stock photo website powered by AI, where image libraries no longer depended on photographers. He even had a name: Stock AI.
Then he stopped to think it through. Getty Images. Shutterstock. These companies had massive legal resources and every incentive to litigate hard against AI-generated image competitors. The opportunity was real, but the litigation risk was serious.
He pivoted.
AI professional headshots.
The logic was clear: every working professional needs a good LinkedIn photo, but booking a photographer takes time, costs money, and requires actually being somewhere with a camera. If AI could turn a handful of phone selfies into studio-quality professional portraits for $30-$50, how many people would pay for that? The answer was: a lot.
He started building.
30 hours to launch, ahead of everyone
Danny understood the idea wouldn't stay secret. Stable Diffusion was public, and thousands of developers would eventually think of similar applications. He gave himself a hard constraint: idea to live product in 30 hours.
The product logic was simple: users upload 10-20 selfies, an AI model fine-tunes on those photos, then generates 100-200 professional-style portraits in various settings and backgrounds. Users download their favorites.
He named it HeadshotPro and flipped on payments on March 16, 2023.
He posted about it on Twitter. The response was immediate — shares, engagement, and purchases. Within two weeks: $100,000 in revenue.
When he made his numbers public, people assumed he was exaggerating.
One year in: $300K MRR
HeadshotPro didn't fade. It was solving a permanent problem: everyone needs a good photo when changing jobs, getting promoted, or updating their professional profile. AI dropped the cost from a $300-500 photographer to a $29-49 package.
Twelve months after launch, HeadshotPro was generating $300,000 per month in recurring revenue. Annualized, that's over $3.6 million. Danny kept operating with zero employees, grouping all his AI products under the Postcrafts brand.
The complete arc: web marketer to "built my own design library because I couldn't find one," to "sold an AI writing tool for $1M," to "built a $300K/month product in 30 hours while bored in Bali."
Source: unite.ai/danny-postma-founder-of-headshotpro-interview-series
Thinking
Danny Postma's case demonstrates a principle that holds across all three of his products: your own genuine pain is the most reliable product research. Landingfolio existed because he couldn't find design references. Headlime existed because writing marketing copy was slow. HeadshotPro existed because he realized everyone needs a professional photo and few people have a good one. This is a method, not luck.
The second thing worth examining: he's fast, but not reckless. Stock AI — he thought it through, identified the legal exposure, changed direction. HeadshotPro — he thought it through, saw the market size, moved in 30 hours. Speed and judgment together. Speed without judgment produces expensive pivots; judgment without speed produces missed windows.
And a third thing: the Headlime exit gave him psychological runway, not just financial runway. When he built HeadshotPro, he wasn't building to survive. He was building from curiosity and conviction, with the patience to launch something unpolished and the confidence to iterate. That mental state produces different work than building under survival pressure.
Action
Validate product direction with your own real needs: What do you repeatedly need in your work or daily life that has no good solution? That recurring annoyance is probably a business waiting to be extracted.
When a new technology appears, ask "what expensive old problem does this make cheap?": Danny didn't see Stable Diffusion and think "cool technology." He immediately thought "what does this make cheaper than before?" Professional photos was his answer. What's yours?
The 30-hour principle — minimize the cost of testing an idea: HeadshotPro went from idea to live product in 30 hours. Can you put a working, paid version of your idea in front of customers within 3 days? If not, ask yourself why the timeline is longer — and whether it needs to be.
Run the legal/structural analysis before you commit to building: He killed Stock AI before writing a single line of code because he thought through the litigation risk. HeadshotPro survived the same analysis. Fast is good. Fast in the wrong direction is expensive.