Twelve weeks. A real product, live on the internet. You do not quit your job.
The first cohort of the Solopreneur Accelerator. Twenty seats, for employed professionals who have carried an idea for years and were told they would need a developer.
Cohort starts 11 January 2027
Twelve weeks, fully online. One live 90-minute session each week, recorded. A peer group of four or five founders you meet with weekly. An async community between sessions. Guest sessions from people who have done specific parts of this.
You arrive with expertise and an idea. You leave with a validated idea, a working product deployed on the internet, a pitch deck, and a 90-day plan for what happens next.
It is built around a full-time job. Ten to twelve hours a week.
If that is not available to you, this is not the right twelve weeks and you should not apply.
Building software used to cost 50,000€ or a co-founder. That was a real price, and declining to pay it for an unproven idea was the correct decision.
That price collapsed. The build is now the cheap part, which means the expensive part is the thing you already have: twenty years of knowing exactly where an industry is broken.
Read the full argumentCheap to build. Funded by the salary you already earn. Defended by expertise nobody can prompt their way to.
Every week runs on the same five moves.
Not a philosophy. The order you do things in, so you never build something nobody wanted.
Prove people want it before you build it. Evidence, not enthusiasm.
Ship the smallest product that solves one real problem, with AI doing the lifting.
Charge from day one. Revenue is the only validation that cannot be faked.
Build the audience while you build the product, not after.
Let the numbers, not your ego, decide what gets built next.
The whole method, printed.
Hype conceals the method. One 90-minute live session a week, and every week ends with something that exists.
Backed by evidence from real potential customers, not encouragement from friends.
Deployed, live, usable by a stranger. Not a mockup, not a Figma file.
The story of what you built and why it matters.
Exactly what you do the day after the cohort ends.
Not a certificate. Not a folder of notes. Not a recording library you will not watch.
Some weeks nothing will work. You will spend a Tuesday evening on something that should take twenty minutes, and go to bed with it still broken.
There will be a week — usually five or six — where the gap between what you imagined and what you have built is wide enough to be genuinely discouraging.
You will have to talk to strangers about your idea before it is ready, and some of them will not want it. That is the point of doing it in week four rather than year two, but it does not feel good either way.
The people who do not finish are almost never the least technical. They are the ones who could not protect five hours a week.
Do not apply if any of these are you.
One 90-minute live session, and the rest in pieces you choose. In practice, for most people: the live session, one long weekend block, and about forty minutes on two or three weeknights.
Every session is recorded, your peer group meets independently of it, and the async community runs between sessions.
Missing one week does not put you behind. Missing four does.
Most people reading this share finances with someone. This is a real conversation at home, not a click.
Everything you would need to have it properly is in the facts block . If it helps, the honest framing is this: twelve weeks and a fixed cost to find out whether the idea you have been carrying is real. Not a bet on it being real.
What you are buying is an answer, not a lottery ticket.
What that is next to the alternatives you have actually priced:
No countdown, no bonus expiring at midnight, no struck-through price that was never charged. The price is the price. The number is not on this page. Apply and you will have it within 72 hours, with the payment options, whether or not you are offered a seat. No call needed to find out.
You have the 14-day withdrawal right EU law gives you. I am extending it. Attend the first two live sessions, do the first week's work, and if you decide this is not for you, email me before the end of week two and I refund you in full.
After that, no refunds — because from week three the only variable left is whether you do the work, and that one is yours.
I am not going to guarantee that you will ship a product. Anyone who promises you an outcome they cannot control is telling you something else about how they operate.
This is cohort one. There are no alumni, and I am not going to invent any.
Twenty seats, and I am in the room for all of them. That ratio does not survive cohort three.
The curriculum bends around the twenty people in it. Later cohorts get the version your questions produced.
This price. It goes up, and you keep this one.
What it costs you: you are the one finding the rough edges.
Luis Gonçalves
Twenty years in digital product development. Led the agile transformation at Nokia. Teaches at Porto Business School.
A venture that lost about 250,000€ — which is where the method comes from.
Then the platform this runs on: built alone, directing AI. You are reading this on it.
A short written application. Fifteen minutes, not two.
I read every one. You hear back within 72 hours, either way.
A short call if it is a fit both ways.
Twenty seats. If the cohort fills, you go on the list for the next one and keep this price.
Domain expertise you have earned, a problem you have seen up close, and five protected hours a week. Not technical ability. Not a finished idea.
You need expertise and a problem you have seen up close. The idea gets narrowed in week two — arriving with a half-formed one is normal.
No. The entire premise is building without being a developer. You direct the build in plain English and learn to check the output.
Ten to twelve hours a week. Every session is recorded, your peer group meets independently, and the async community runs between sessions. Missing one week does not put you behind. Missing four does.
Read your contract before you apply, and if it is ambiguous, get it looked at. Build in your own time, on your own hardware, in a domain unrelated to your employer. This is your call to make, not mine.
This is the most credible objection in the market, so here is the honest answer: you will be able to change what you built, because you built it and you keep the method. But if you stop touching it for a year, you will have to relearn it. That is true of any codebase, including ones written by teams.
Entirely. No equity, no licence, no revenue share. The account, the code, the domain and the customers are yours.
You leave with a 90-day plan. If you want a room to keep going in, that is The Room — the membership for people who have shipped. It is optional.
Yes. In full, or across three or six months — the first when you accept your seat, then one a month. The exact figures come with your application response.
The tools will change. The method — validate, scope, build, ship, sell — does not. That is deliberately what the twelve weeks teaches.
You can. What you cannot get there is a deadline, twenty people at your exact stage, and someone telling you the truth about your idea in week three rather than year two.
The twenty years of knowing where an industry is broken is the input that cannot be bought or prompted. The code is now the cheap part. You have the expensive part.
You keep everything you built and the recordings. Nobody chases you. The people who do not finish are almost always the ones who could not protect the hours.
People who want it built for them, people who want to raise and hire, and people who cannot protect five hours a week. See the section above.
A validated idea, a working product deployed on the internet, a pitch deck, and a 90-day plan.
You already know what the next five years look like if nothing changes, because you have just lived them.
The idea will not stop costing you attention until you find out whether it is real.
Five more years is five more versions of this December. The same folder, opened a little less often.
What is left is twelve weeks and whether you protect the hours.
You have carried this for years. The reasons not to build it were real, and they have expired one by one — the developer, the co-founder, the hundred thousand, the quitting.
Twenty seats. Applications close 4 December 2026.
Not sure? Talk to us first