We build what itches
No roadmap, no thesis, no TAM slide. If a problem keeps bugging us for a week, it earns a week on the bench. That’s the whole selection process.
A workshop, not a startup
Software objects, built on one bench and shipped whether they're ready to be famous or not. The extra g is load-bearing.
CATALOG STATUS — objects shipped: 11 · pitch decks produced: 0
Field notes
No roadmap, no thesis, no TAM slide. If a problem keeps bugging us for a week, it earns a week on the bench. That’s the whole selection process.
AI handles the scaffolding, boilerplate, and busywork. A human keeps the taste, makes the calls, and signs the finished object.
Every thingg ships with honest notes — what worked, what broke, what we’d do differently. If it flops, the flop gets written up too.
The catalog · newest first
The whiteboard, transcribed
An agent that wrangles the tasks, outreach, and research a solo founder is currently doing at midnight. Might earn a catalog number soon.
Brand, landing page, and launch checklist for a niche idea, generated while the coffee is still hot. Sketched, not started.
Little automations that trawl communities and job boards for problems nobody is building for. Currently a pile of bookmarks.
The best objects in the catalog started as someone else’s complaint. If something bugs you every single week, tell us about it below.
Correspondence
It’s a workshop. Studios raise funds and write theses; we make objects and number them. If one of the objects turns into a company, great — that’s the object’s problem.
An idea has to itch for at least a week and be buildable in about one. That filter is stricter than it sounds — most ideas fail the boredom test before they fail the feasibility one.
Yes. Bring domain expertise, a gnarly problem, or a dare. The bench has room for guests, and co-built objects still get a number.
There is exactly one person at this bench — Alex — and every message lands there. Questions, ideas, dares, collaborations: all read.
Replies written by hand, not by the robots.