I joined ProAI when it was the founder and me. I built the platform that served 4,000+ students, and I ran the launches that paid for it.
ProAI had AI education material that was genuinely worth paying for, and no way to deliver it at scale. The courses sat in separate places, so a student had to figure out on their own what came after what. The experience was fragmented and there was no tracking anyone could act on. We could not say which student was stuck, or where, or why they stopped.
That matters more in this category than in most. People buy an online course and then do not finish it. Low completion is the normal outcome, and a platform that ignores it inherits it.
The other half of the problem was who was thinking about what. Nobody was holding product and growth at the same time. The launches got written by one side of the room, the platform got built by the other, and the two drifted apart every time we shipped.
The thing we actually wanted was simple to say and hard to build. One platform where a student runs a structured AI master's journey from start to finish, instead of buying isolated courses and hoping they add up to something.
I set the roadmap, made the architecture calls with the engineers and ran the sprints. I also wrote the go to market for every launch. That was on purpose. When the person deciding what gets built is the same person who has to sell it in six weeks, you stop building features nobody asked for.
We were on a custom solution that did what we told it and told us nothing back. I led the move to LearnDash on WordPress, and that gave us analytics and user segmentation for the first time. From that point we could see where students dropped, which cohorts bought what, and who to talk to next. Before it, every decision was an opinion.
The hardest call of the whole two years was short-term launch demands against long-term scalability. Every launch wanted something now, and every shortcut we took to give it was a bill we would pay later. The fix was a release roadmap where the launch features shipped on the launch calendar while the refactors ran in parallel, in their own lane, instead of waiting for a quiet quarter that was never going to come.
We went through an accelerator and raised a pre-seed round, and the team grew past 10 people. I hired and onboarded as we went. Growing a team while you are also shipping is its own job, and nobody warns you that the roadmap has to change shape when the number of people working on it doubles.
Nothing here came from posting and hoping. Every launch had a two-week runway before the offer was ever mentioned. Email sequences that taught something first. Social proof from the students already inside, because people buy education from people who look like them. Free webinars where we gave away the actual method, and the offer came at the end for the people who wanted it done with us instead of alone.
Under that, the funnels were segmented. A person who watched the whole webinar did not get the same email as someone who opened one link, and retargeting picked up the ones who left. Then the discovery call did the rest. 60% of those calls became paid students, because by the time someone got on a call with us they already knew what they were buying.
The part most people skip is what happened after the payment. Onboarding was personalized, one student at a time, and that held a 90% completion rate on the master. In a category where most people never finish, that number is the whole business. Students who finish stay, they talk, and they are the social proof for the next launch.
The channel here was email, webinars and paid social. Not search. But the sequence underneath is the one I still run: find out what buyers are already trying to learn, look at what the rest of the category is selling and where it is weak, build the thing that answers it, put it where those people already are, then measure and go again.
That sequence is the OCAVI framework I install for clients today, with Google and AI answers as the channel instead of Meta and Mailchimp. ProAI is the reason I know it moves revenue and not only traffic. The numbers on this page are money, not pageviews.
The first Black Friday launch did €85K in 7 days. That was the moment the platform and the launch machine were finally the same system instead of two teams pulling on the same week.
Over the two years the platform served 4,000+ students and the company reached €500K+ ARR, with a team that went from the founder and me to more than 10 people, an accelerator behind us and a pre-seed round closed.
I left with the thing I use every day now. I had seen a product and a go to market built by the same hands, and I had the receipts on what that does to conversion.
The build and the way it gets sold are one decision. Split them across two teams and you get a product nobody knows how to describe.
A stack that cannot segment your users keeps you guessing. Moving to LearnDash was the day we could stop arguing about what students were doing and go look at it.
Two weeks of teaching, proof and answering questions, then the offer. A launch that starts on launch day is a launch to nobody.
The customer who finishes is the one who renews and the one who refers. Personalized onboarding held completion at 90%, and that fed the next launch more than any ad did.
Waiting for a calm quarter to fix the foundations means never fixing them. Same reason I fix the technical foundations of a site before publishing a single page for it.
ProAI worked because the research came first and the building came second. That is the order I install in 90 days: where your buyers are searching, where your competitors already win, what you can realistically take, then the pages that get you found and cited. If that sounds like your problem, let's find out on a call.
See if we're a fitRun the free visibility check →