A week in a one-person venture studio

People assume a venture studio is a building full of people. Ours is mostly one operator, a lot of cloud, and a stack of AI agents that turn slow expert work into software. The portfolio on the home page isn’t a roadmap — those are things running today. The honest question I get asked is: how do you ship across that many things at once?

The answer is unromantic. You pick the one excellent version of each thing and you let the rest wait. A single great outreach email beats five mediocre ones. One screener that actually works beats a dashboard nobody opens. Breadth is only possible because the depth bar is “ship the one thing that’s genuinely good, then move.”

What a typical week touches

A defense-innovation intelligence platform used by an allied innovation accelerator. A home-inspection app that turns a phone walkthrough into a defensible report. A site-intelligence engine scoring land for wind projects. A FEMA grant-closeout system reconciling thousands of invoices. And — because the studio is also just a person with a life — a direct-booking site for a lake-house retreat and a coaching brand for a friend in Athens.

They look unrelated. Under the hood they share more than you’d think: the same deploy patterns, the same agent scaffolding, the same hard-won rules about keeping each client’s data walled off from the next. Build the muscle once; point it at the next problem.

The part that doesn’t scale

Taste doesn’t scale, and that’s the point. The agents can research, draft, deploy, and remember. What they can’t do is decide which of the ten possible things is worth building this week. That’s still the job — and it’s the most fun part.

If any of this is useful to you — a tool you wish existed, a workflow that’s eating your week — that’s exactly the kind of thing we like to pick up.