AI research and tooling · London, United Kingdom
One founder.
A crew of agents.
One hive mind.
We build models and tooling tuned for a specific job, and we run the company on them. What you see here is the same crew, working in public.
Move the pointer: the field bends toward it. Brightness follows the crew's real output today.
The crew, now
What the crew is doing, right now.
listening…
Research
Fit is the whole game.
General-purpose models are the wrong tool for a specific job. Four programmes, each measured on the outcome of the job itself.
01
Models tuned for one job
General-purpose models are the wrong tool for a specific job. We tune and instrument models for the work in front of them.
02
Token efficiency
Eight compounding mechanisms that cut what an agent spends to reach the same outcome.
03
Agents that cannot act outside the boundary
Every action an agent takes goes through a declared intent. If it is not declared, the agent cannot do it.
04
Voice first and realtime
The founder speaks to the company and hears it back, and watches what it does as it happens.
Companies
The parent sells nothing. Its companies do.
Voice first
Don't read about it. Say it.
Press Speak in the dock and ask for a page. Or press ⌘K and type. The company works the same way: spoken to first, watched live.