Productivity · AI Business
He Couldn’t Code When He Started. Now One Guy Runs 20 AI Products From Bali.
Danny Postma never trained a model or invented an algorithm. He just got very good at one thing — combining tools that already exist. That skill is now worth a reported $300K a month.
In 2023, Danny Postma shipped a product called HeadshotPro. It does something almost embarrassingly simple: you upload a few selfies, and it generates professional-looking headshots. He built it in roughly 30 days. Within two weeks of launch, it had reportedly pulled in $100,000.
It now does a reported $300,000 a month and has generated something like 18 million AI headshots for nearly 197,000 customers.
Here’s the part that should make you sit up: Danny Postma is not an AI researcher. He’s a marketer from the Netherlands who, for years, couldn’t write a line of code. And that turns out to be the whole point.
The era of building the tech is over
For most of software history, the moat was can you build it. The engineers won. If you couldn’t code, you needed someone who could, and you split everything with them.
AI quietly flipped that. When powerful image generation, text generation, and voice models are all sitting one API call away, building the technology is no longer the hard part. The hard part — the rare part — is knowing what to build, for whom, and how to hand it to them so smoothly they’ll pay for it.
The question changed from “how do I make this?” to “what do I make, who is it for, and how do I deliver it?” Postma’s entire career is a case study in the second question.
A timeline built on combining, not inventing
2015, age 21. He’s a freelance web marketer specializing in conversion optimization. He keeps running into the same problem: the landing pages his clients need just aren’t good enough. He can’t code, so he doesn’t build a tool — he builds Landingfolio, a gallery of landing pages, using nothing but WordPress. A marketer’s eye, off-the-shelf tools, zero engineering.
2020. He finally teaches himself to code at 25, and ships Headlime, an AI copywriting tool. He’s one of the early indie developers to wire GPT-3 into a real production product. The result: a lifetime deal that reportedly did $60,000 in its first week. By March 2021, sitting at $20,000 in monthly recurring revenue, he exits the business for a reported $1,000,000.
2023. HeadshotPro, the 30-day build that became a reported $300K-a-month machine.
Today he runs a one-person studio called Postcrafts, operating around 20 AI products at once — HairstyleAI, PhotoShed, Deep Agency, and more — reportedly mostly solo, from Bali. One person, doing what a mid-sized company would take years to do, at many times the speed.
The numbers that explain why
A lean AI product like this can reportedly run at a 60–80% operating margin, versus the 10–20% a typical small business lives on — roughly 4 to 8 times the efficiency. The reason is brutally simple: there’s no payroll. The whole “team” is a software stack that costs a few thousand dollars a year to run.
The actual formula (it’s less sexy than you’d hope)
There’s no secret model architecture here. Strip away the story and Postma’s playbook is three moves anyone can copy:
Find the demand first. He didn’t start with a clever AI and look for a use. He found a specific group of people with a specific, paid pain — people who want a good headshot but won’t book a photographer — and then went looking for the tools to solve it.
Combine existing APIs into one clean experience. Image generation here, text generation there, glued into a product that feels effortless to the customer. The genius is in the packaging, not the plumbing.
Automate the boring operations away. No-code and automation tools handle the repetitive work, which is exactly how one person can keep twenty products alive without drowning.
Why this matters right now
For the next decade, the winners among solo builders won’t be the best engineers. They’ll be the people who read market demand fastest and combine existing AI cleverly. Curation is becoming the core skill — the taste to spot a real problem and the judgment to wire together the right tools.
That’s genuinely good news if you’ve ever felt locked out because you’re “not technical.” The thing that used to disqualify you matters less every month.
Don’t romanticize it
Postma is one standout outcome, not a guarantee — and notice he did eventually learn to code, and spent years building taste and an audience first. “Combine some APIs” is the easy sentence; finding a pain people will actually pay for is still the hard, unglamorous work. The tools got cheap. The judgment didn’t.
Who this is really for
| ✅ This fits you if… | ⚠️ Think twice if… |
|---|---|
| You can spot real problems people pay to solve | You want passive income with no marketing effort |
| You’re happy combining tools instead of inventing them | You’re waiting for one perfect, original idea |
TL;DR
Danny Postma went from a marketer who couldn’t code to a solo founder running ~20 AI products at a reported $300K a month — not by building better AI, but by finding demand and combining tools that already existed. In the AI era, curation beats construction.
Steal the playbook this week
☐ Pick one group of people whose pain you understand
☐ Find an existing AI API that could solve it
☐ Wrap it in the simplest possible product
☐ Ship a rough version and charge for it
The barrier was never the technology — not really. It was knowing what to make and having the nerve to ship it. That part has always been available to you.
