I build practical tools that solve actual problems. No corporate jargon, no roadmap theatre, no promises about the future of anything — just software that behaves the way it should have in the first place.
I am not chasing trends. I build the tools I want to exist, the way I want to use them, and then I leave the source where anyone can take it.
From saving five seconds opening a Medium article, to putting several AI models behind one window, to a desktop app that turned out better than the official one — the through-line is the same. Remove the friction. Do not add any.
Build stuff that makes sense. That is the whole strategy document.
- 01
Open source first
Every project here is open. Transparent, auditable code that anyone can read, fork or argue with. Open source is not a marketing word — it is what keeps the work honest.
- 02
Built for the person using it
Built by a developer, for developers. If a tool needs a tutorial before it makes sense, it is not finished yet.
- 03
New where it helps
AI, automation, whatever is useful — used where it solves the problem, not where it makes the README sound impressive.
- 04
Works, looks right, breaks less
Three things, in that order. A tool that works but is unpleasant gets abandoned. A tool that is pretty and unreliable gets uninstalled.
- 2023
Inulute starts
Solo dev mode on. One person, a text editor, and a list of small annoyances.
- 2023
Medium Unlocker ships
The first tool that other people actually used. Paywalls, politely ignored.
- 2024
Simplexity AI Desktop
A native shell that outgrew the official one. SourceForge named it Community Choice.
- 2024
AI Gate launches
One gateway to every model, because six browser tabs is not a workflow.
- 2025
A community shows up
People shipping, testing, and breaking things — politely, mostly.
No office, no team page, no time zone that matters much. If you use one of these tools and something is broken, the issue tracker is the fastest way to reach me.