Why I build twenty companies at once
Conventional advice says focus on one thing. I'm doing the opposite, and it only works because of one decision I made about the foundation underneath all of them.
The standard advice is to focus on one thing until it works. I run a portfolio of twenty ventures at once. That sounds reckless. It isn't, but only because of a single decision about what sits underneath them.
The case against focus
Focus is good advice when the cost of starting the next thing is high. For most of software history it was. Every product meant a new team, a new stack, a new operations burden, a new everything. Starting a second company while the first wasn't done was how you killed both.
AI changed the slope of that curve. The marginal cost of standing up a credible product, with real backend, real automation, and real polish, has collapsed. When the next venture is cheap to start, "focus on one" stops being a law of physics and becomes a preference.
I have a different preference: build a portfolio, not a product.
The one thing that makes it work
Twenty companies is chaos if each one is its own island. The thing that turns it into a portfolio instead of a pile is a shared, governed foundation every venture sits on:
- the same self-hosted infrastructure
- the same deploy and operations playbook
- the same AI-native patterns for agents, content, and automation
- the same guardrails so nothing ships that lies or breaks
Build that substrate once, and each new venture inherits it. The twentieth company is cheaper to start than the second one was. That's the entire bet.
What's actually shipping
This isn't theory. The portfolio is real and at different stages:
| Stage | Examples |
|---|---|
| Live | PyroSync, self-hosted scraping infrastructure |
| Launch-candidate | Agency Script (the flagship), CaseSolo |
| In development | Girard AI, ReplyType, CoreReflex, HostSSH, and a dozen more |
Some will work. Some won't. The portfolio is the hedge, and the shared foundation is what makes running it survivable for one person.
The honest tradeoff
A portfolio buys optionality and pays for it in depth. No single venture gets the undivided attention a focused founder would give it. I'm betting that a great foundation plus broad coverage beats one perfectly-tended product, for this operator, in this moment, with these tools.
Ask me again in a year.