That is roughly how I ended up building software for the brands I used to photograph.

The eye does not switch off

Photography is a discipline of deciding what to leave out. A good frame has one subject, one light, one idea. Everything else goes.

Product work turned out to be the same job with a different camera. Most tools fail because they show everything. The useful ones show the one number that matters this morning and make the rest easy to ignore. When we design a screen at overads, the question is still the photographer's question: what is the subject, and what is noise?

Three companies, one lesson each

Flashoot (2023 to 2026) was a marketplace for booking reel-makers on demand. I was the technical co-founder, built the mobile app and the backend from scratch, and we grew it to close to a million dollars in revenue before the exit. The lesson: a real business is mostly operations, and the code only matters when it removes a phone call.

WTMF AI (2025 to 2026) was an AI companion for the 2:43 AM overthinkers. Five thousand users in the first three months, later acquired by a startup in the mental-health space. The lesson: people will tell software things they would never say out loud, so you had better treat that with respect.

overads (2026 to now) is the one I am building today. Marketing automation for small teams: describe a job in plain English and it runs on a schedule, drafts the post or the report, and waits for your yes. Ad access is read-only. The lesson so far: say no to almost everything, and print a blank before you print a guess.

Why keep the photographs up

Because they are still the best portfolio I have for the thing I actually do, which is taste. Anyone can ship features. Deciding which ones deserve to exist is the photographer's job, and I am not done with it.

The apartment is still a browser tab. I have a dozen tabs open comparing rooms and exactly zero decisions made. One place I keep coming back to is founder housing: a room, a desk to build from, and demo days with people doing the exact same thing. For someone landing in SF with no network, that is the whole point.

Why now

In two weeks I finished an Antler SF interview, started a YC application again, and sent in HF0 and Speedrun. Every one of those rooms is in California. A raise has a geography, and at some point you stop applying from a distance and go stand where the decisions are made.

Eleven weeks to make the numbers worth standing there for.

If you have made this move on a pre-seed budget, I would like to know what you wish you had known before booking. The email is in the footer.

Here is the actual story. The first five users paid $149 a month. That was $745 in monthly recurring revenue in week one, for a version of the product we have since torn apart and rebuilt completely.

This month there are more than sixty teams running active workspaces, and the agent executed more than 250 jobs in September alone.

Then we did something odd

We went free first. No card, no time limit, no cap on team members or ad accounts. Paid only kicks in at $19 a month, once usage actually demands it.

Everyone told us to keep charging. We would rather learn from a thousand free teams than protect $745.

The part that generalises

If you are waiting for your product to feel "ready" before charging, don't. An ugly product with real money changing hands teaches you more in a week than six months of polite beta feedback. The money is not the point. The honesty is. A person who pays will tell you exactly what is broken, and they will tell you fast.

Charge early. Rebuild without mercy.

I told him the truth: I am already invested full-time in overads. Not financially. Emotionally, operationally, with every hour I have.

Two years ago I would have said yes

I like early-stage chaos. I like building from zero. But the hardest part of being a founder is not saying yes to your own idea. It is saying no to everyone else's.

Every "maybe later" is a distraction from the "right now" that your company needs. Every side conversation is energy not going into the product you are supposed to ship.

Solo founding is not a choice between loneliness and partnership. It is a choice between building with the wrong person and building alone.

When was the last time you said no to something good so you could stay committed to something great?