How we work

A short path from idea to a system in daily use.

Evidence at every step, and nothing you can't see, measure or own.

Two engineers working side by side at one desk, one pointing at the screen while the other types.

Every prototype ends with

An evaluation on your own cases, and a clear call: build, change course, or stop.

  1. A first conversation

    You tell us about the work you want to change. We say plainly whether AI is the right tool for it.

  2. Discovery

    We study the work, the data and the constraints, and agree what success looks like in numbers. [Typical length of a discovery.]

  3. Prototype

    A working version on your real data, with an evaluation you can read. Then you decide whether to build.

  4. Build

    Production software with security, monitoring and documentation, built with your team rather than handed over the wall.

  5. Run and improve

    We watch quality, cost and speed in use, improve what drifts, and hand over as much as you want to run yourselves.

What we'll ask of you

  • One person on your side who owns the outcome
  • Time with the people who do the work today
  • Sample data, or a safe way to reach it
  • Honest feedback on every prototype

How we handle your data

  • Used only for your project
  • Seen only by the people working on it
  • [Where data is stored and processed]
  • Returned or deleted when the work ends, as the contract sets out

Ways to work with us

Talk to us

Tell us what you're working on.

A few lines are enough. A person reads every message and replies.

We'll reply to this address.

The work you want AI to help with, in a few lines.