Free website and AI help, one project a month
One project a month.No catch, no invoice.
Every month Bottleneck Labs builds one project for free: a small website, one automation, or a small AI system over documents you already have. It is for nonprofits, community organizations, public interest projects and one-person operations that cannot pay a consultancy. It is genuinely free, not a trial, not a discount and not a lead magnet. There is no invoice at the end, and one project is chosen each month.
The program runs out of Toronto, remotely across Canada. You own the code, the accounts and the keys when it is done, the same as a paying client. The reason it exists is simple: this work is cheap for me and expensive for you, and it is worth far more to the people who get it than it costs me to do.
What gets built, and what does not.
What gets built
- A small website that loads fast, reads well on a phone, and you can edit yourself afterwards.
- One automation that removes a job somebody is doing by hand every week: a form that files itself, a report that assembles itself, a list that stays current without anyone retyping it.
- A small AI system over documents you already have, so people can ask a question in plain language instead of hunting through a shared drive.
- A blunt read on a tool you are about to buy, including the answer that you should buy it and not hire anyone.
What does not
- Anything that needs more than about two weeks of my time. One month, one project, and the month has paid work in it too.
- A platform, a marketplace, a mobile app, or a system with accounts, payments and roles. Those are real projects and they need a real budget.
- Ongoing support. I hand over the code, the accounts and the keys, and then it is yours. I will answer questions, but I am not on call.
- Anything that needs a subscription you cannot afford after I leave. If the running cost is more than a few dollars a month, it is the wrong build for this program.
- A rescue of a project somebody else abandoned mid-build. Untangling that takes longer than starting over, and it rarely ends well for free.
None of that list is a negotiating position. A project that needs more than the ceiling allows does not become possible because the cause is good, and pretending otherwise would waste a month you do not have.
Don't write me a brief.
Ramble. Thirty seconds of voice note beats a paragraph you spent an hour on. Making sense of it is my job.
Applications that arrive by the last day of the month are considered for the next month's build.
One a month, and everyone hears back.
One project a month is the whole capacity. Most months there are more good applications than slots, which means a no is usually about timing and fit rather than about the work being unworthy.
Every application is read by me. Nothing is scored by a model and nothing is discarded unread. You hear back either way, including when the answer is no, and a no comes with the reason plus a pointer to something cheaper or simpler if I have one.
A no in one month is not a permanent no. Apply again when the problem is still there.
What makes an application strong.
- One thing that is stuck, named in a sentence. Not a roadmap.
- Who it hurts, and how often. A weekly grind beats a hypothetical.
- What you have already tried, including the tool that did not work.
- Whether anyone on your side can keep it running after I hand it over.
What people ask before they apply.
Is the pro bono work really free?
Yes. No invoice, no deposit, no hours billed later, no equity, no revenue share. You own the code, the accounts and the keys when it is done, the same as a paying client. The only thing I ask for is permission to write about the work, and you can say no to that and still get the project.
What is the catch?
The catch is scope and scarcity, not money. One project a month, roughly two weeks of my time, and the ceiling is real: a small site, one automation, or a small AI system over documents you already have. If your problem is bigger than that, this program is the wrong fit and I will say so rather than start something I cannot finish.
Who gets picked?
The application where one specific thing is stuck, the people it hurts are named, and somebody on your side can keep it running after I hand it over. Small nonprofits, community organizations, public interest projects and one-person operations get priority over funded companies. I do not pick on how polished the application reads.
What happens if I am not picked?
You hear back either way. Every application is read by me, not filtered by a scoring model, and a no comes with the reason and, where I have one, a pointer to something cheaper or simpler that would solve it. A no in one month does not disqualify you from applying again.
Can I pay you instead?
You can, but that is a different conversation and it belongs on the services pages, not here. Paying does not buy a place in this program and does not make an application move faster. The pro bono slot exists for people who cannot pay a consultancy, so it stays that way.