12 apps in production

Today the portfolio crossed 12 apps in production.

The twelve ship across five disciplines: mobile apps on Flutter and Kotlin, web tools on Next.js and FastAPI, automation pipelines on Python, games on Dart and Compose, and a handful of AI utilities that prefer bring-your-own-key over rented cloud inference. We started in 2024, and the count went from zero to 12 over roughly two years of concentrated solo work plus the occasional contract contributor. Most of the shipping happened in six-week cycles, one app per cycle, with a ramp-up month and a polish week at the end of each.

What made 12 possible is the decision to ship narrow. Every app we run does one thing well and stops. The Mieter App checks Mietpreis and flags Nebenkosten; it does not try to also be a general legal assistant. SEOMAX tracks ranks and clusters keywords; it does not try to be a full CRM. The Poker Platform is a card-game engine with admin tooling; it is not a social network. Narrow scope is the only way a one-person studio can credibly ship twelve things in 24 months.

Crossing 12 changes the operational picture. The next app we build will be the first one where we know from day one that onboarding, support triage, and release rhythm matter as much as the feature list. We are adding a support inbox router in May that threads device + version + app through a single triage queue so the twelve-apps-one-developer math still works at 15 or 20 apps.

Thanks to the Flutter, Dart, Postgres, FastAPI, and Python ecosystems that did most of the heavy lifting. Thanks to the hundred-and-change beta testers who told us which of the 12 apps deserved more attention, and thanks to the twelve-bill photographer who handed us the Betriebskosten OCR corpus we are building 1.5 on.