This site is a search query. I write so people who build things can find me. That is the whole reason this page exists.
Most of what gets written about a fifteen-year-old who ships things is written by someone who wasn't there. So here is the primary source. Seven things between ages 8 and 15 — what each one was, what it taught me, and where it went wrong. Two of them made money. Two of them went nowhere. One of them I sold.
Where I don't have a number in front of me, I've left it out rather than round it up.
Age 8 — first code
Code Ninjas
Code Ninjas is a coding class for kids. You work through levels at your own pace and you build small games as you go. I signed up at 8.
The part that stuck wasn't the syntax. It was the loop: you have an idea, you type for a while, and then a thing exists that runs without you. At 8 that is close to magic, and I don't think the feeling has actually changed since. Every project on this page is a version of that same loop with higher stakes.
Most kids cycle out after a year. I never stopped.
Age 10 — first business
Pocket Board
A board game small enough to fit in a pocket. I made it, and then I sold it door to door.
Making it was the easy half. The other half was standing on a stranger's doorstep at 10 years old explaining why they should give me money for a game they had never heard of. A lot of no's before the first yes.
What I actually learned there wasn't sales technique. It was that a no costs you nothing and you find that out within about twenty doors. Every business decision I've made since has been cheaper than it felt, and that's where I learned it.
Age 11 — second place
The hackathon
I placed second in a hackathon much bigger than I was.
The useful thing about a room like that is that nobody checks your age at the door. There is a demo or there isn't. That's the entire evaluation. Rooms like that let you in if you just build, and once you've been in one you stop treating the other rooms as closed.
Second place also meant I lost. That's worth saying out loud, because most of this page is people describing the parts that went well.
Age 12 — filed away
Too early
I tried to build legal tech. The AI wasn't ready. I wrote it down and waited.
This is the chapter I'd point at if you only had time for one. It looks like the year nothing happened, and it's the reason the next few worked.
The idea was right and the tools underneath it weren't. Legal paperwork is structured, repetitive, high-stakes and miserable — a perfect target — but the models at the time couldn't be trusted to read a document and get the details right, and in that domain getting the details nearly right is the same as getting them wrong. So I stopped. I didn't pivot it into something worse. I wrote down what would have to be true for it to work, and I left it alone.
Two years later the thing that had to be true became true, and I already knew exactly what to build.
Almost nobody shelves an idea on purpose. You either force it and burn a year, or you forget it. Writing it down is the only part that's hard, and it's the reason ProbateDesk took months instead of years.
Age 13 — real revenue
Supremely Rare
Rare snack drops. I bought things in the US that you couldn't get in Canada and sold them here. Then I found out there were countries that couldn't get them at all, and sold there too.
This was the first thing I did with real revenue, and it was the least technical thing on this list. Almost none of the work was code. It was sourcing, ads, margins, customs and guessing demand — and getting any one of those wrong ate the margin on everything else.
Two things I actually took from it. First, a margin is not a number you set, it's what survives after shipping, fees, returns and the units you bought that nobody wanted. Second, scarcity is a real product feature and it is also a real trap: the thing that makes people want it is the same thing that makes it impossible to restock.
Everything I know about running the business side of Anticipy — pricing, pre-orders, what it costs to actually put a physical object in someone's hands — I learned selling snacks at 13.
Age 14 — sold to a law firm
ProbateDesk
Software that automates British Columbia probate paperwork. I built it at 14. A law firm bought it. It's still running.
Probate is the court process for settling someone's estate after they die, and in BC it runs on a set of forms that have to be filled in consistently and correctly. The work is repetitive, it's exacting, and the cost of a small mistake is a rejected filing and weeks of delay. It is close to the ideal case for software: the rules are written down, the output is structured, and the people doing it by hand hate doing it by hand.
By 14 the models were good enough to do the reading. That was the thing that had to be true, and it had become true. So I opened the notes from two years earlier and built it.
The full account — what it did, what broke, and how a fourteen-year-old ends up selling software to a law firm — is on its own page.
Age 15 — now
Aevoy, then Anticipy
Aevoy was an attempt at something much larger than I could build. I chased AGI and missed.
What came out of it was worth more than what went into it. Along the way I built browser agents — software that drives a real browser the way a person does, clicking and typing through a task from start to finish. And they worked. Not in a demo. They could do real work, unsupervised, on the actual web.
The reason I was even looking at browsers was English class. ChatGPT was banned, and I kept wondering what would happen if an agent could just do the assignment end to end, in a browser, like a person. I wired one together to find out. Then I realized the essay was never the interesting part. The interesting part was that the thing could act.
Which left one question: what tells it when to act?
Not a chat box. You don't open a laptop to say "I'll send that tonight" — you say it to a person, out loud, and then you forget. So the input had to be the room you're already in. That's the pendant: 8 grams of titanium, one LED, no screen. It hears what you agree to do and hands it to an agent on your computer.
What I'm after
Where this goes
I'm entering grade 10 at Mulgrave in West Vancouver. Outside of that, I'm building Anticipy full time and shipping the first units.
The reason I wrote all of this down is narrow. I want to be findable by people who build things — founders who've done hardware before, engineers who've shipped BLE firmware that didn't drop packets, investors who look at fifteen-year-olds without either dismissing them or making a story out of it.
If that's you, email me. I reply to every one: omar@anticipationlabs.com
What I'm working on this month is on the now page, and it's updated monthly.