This is the last of these essays for a while, and it is the one I was most tempted to write dishonestly.
Every founder knows the "first ninety days" post. It is a victory lap — a list of wins, a few flattering numbers, a grateful nod to the team, and not a single word about anything that went wrong. I understand the temptation completely, because I am writing this at the end of our first ninety days of building Simsy AI in public, and I have a list of things I would much rather not mention. But a retrospective that only lists wins teaches the reader nothing and earns no one's trust — least of all the investors and institutions watching to see whether we are the kind of company that tells the truth when the truth is inconvenient. So this is the real accounting: the wins, yes, but also the misses, because building in public means publishing both, or it means nothing at all.
What we said we'd do
Let me start by holding myself to what we actually said. Ninety days ago we made three public commitments: to launch the full product ladder and the founder cohort; to deepen the conversations that would fund the next stage of the company; and to open our first serious institutional pilots. We said these things in front of everyone, on purpose, so that this paragraph would one day be possible. Here is how each of them actually went.
What we shipped
First, the things that simply got built — because outputs are the easiest part to be honest about, in that they either exist or they do not. We shipped 1,000 credentials across the four tiers of our credential ladder — from AI Certified Founder Potential, through AI Certified Innovator, to AI Certified Entrepreneur, the last of which spans eight builders and enablers tracks — and we ran multiple cohorts of our AI-Native Founder Cohort. We published 12 of these Field Notes essays, each one a permanent asset on our own site rather than rented space on someone else's platform. And we launched the newsletter you are reading, which now reaches more than 60,000 people. None of that is the company succeeding; it is the company showing up — a lower bar, but a real one, and more than a lot of plans survive to clear.
What worked
Now the part that went better than I expected. The inbound from institutions outpaced anything we had planned for — serious interest across all five of our ecosystem verticals: education, industry innovation, the startup ecosystem, government and policy, and development and impact. One of those institutional conversations moved faster in days than I thought possible in weeks or months. And on the founder side, applications to the AI-Native Founder Cohort exceeded the cap and forced us into a real acceptance process — which is the good kind of problem to have. The clearest single signal was the raw demand: more than 24 institutional conversations opened, and over 150 qualified applications for a cohort with far fewer seats. I am wary of reading too much into ninety days of data, but that one I will allow myself to be quietly pleased about.
What didn't work — and what we got wrong
And now the part the victory-lap version leaves out. The honest bottleneck this quarter was me. Almost every institutional conversation we opened ran through the founder, which is flattering for exactly as long as it takes you to realise it is also the ceiling. We had built a pipeline that could not move faster than 2–3 people's calendars — and the cost of that is not abstract. Conversations that deserved momentum sat waiting on me, and a few cooled while they waited. I am not going to dress these up. We got the shape of our own sales motion wrong — we let the founder + 2 others be the entire revenue function — and the cost of being wrong was a slower month than it needed to be, and the quiet knowledge that the company cannot scale on one + 2 people's relationships. That is precisely why the next hire we are making is the one that takes selling off my desk. I am telling you this not as performative humility, but because the next person deciding whether to trust us — to enrol, to invest, to sign a pilot — deserves to know that when something does not work here, we say so, in public, with our name on it.
What surprised us
There was also something I did not see coming, and the surprises are usually where the real learning hides. I built this company picturing the independent founder — the person sitting on an idea for two years. The audience that leaned in hardest, though, was institutional: the universities, free zones, and ministries who saw the operating system and immediately asked how it could run inside their own walls. I am still working out what that means for where we point next, but the pull is unmistakable, and it is not where I was looking. That is the honest state of things at ninety days: some of the most important data is the data you do not yet know how to read.
What we're keeping: the number stays public
Here is the part that matters most for everything that comes after. We are going to keep publishing this. The production rate — how many ventures actually launch per thousand people who enter the top of our system — we will keep showing, every month, whether or not it flatters us. This retrospective will run every quarter, with the same four questions: what we shipped, what worked, what did not, and what we got wrong. Not because transparency is a nice word to list on a careers page, but because it is the only thing that makes the rest of it mean anything. A claim you make once is marketing. A number you publish every month, in public, that anyone can watch move, is something much closer to evidence — and we would far rather be judged on the second.
What's next
As for what comes next: the work of the coming quarter is less about saying and more about producing — turning conversations into committed institutions, applications into a cohort that actually ships ventures, and the system into something that visibly compounds. I will spare you the targets, partly because I have argued in these pages that loud public targets are their own kind of dishonesty, and partly because the only target that matters is the one you will be able to see for yourself in the monthly number. Watch that, not my promises.
The startups of the future aren't destined to fail. I have ended every one of these essays with that line, and meant it a little more each time — but I have come to think the sentence has a quieter second half. The startups of the future aren't destined to fail, and the companies built to serve them aren't destined to drift into the comfortable dishonesty that lets everyone claim everything is always going well. We get to choose. We are choosing to show the real numbers, the real misses, and the real learning, for as long as we do this.
This is the last Field Note of the first season. The essays will pause; the work, and the monthly accounting of it, will not. Thank you for reading the honest version — it is the only one worth writing.
— Satish Kumar N · Founder & CEO, Simsy AI · Dubai · Bengaluru · New York
Build. Launch. Scale.