Four stages.

Larger projects repeat the cycle in smaller loops. Smaller ones may need only one or two stages.

  1. 01

    Understand

    We start with the problem rather than the technology.

    • What the system should do, and for whom
    • The data available and any limits on its use
    • How success will be measured
    • A short written proposal with scope and timeline
  2. 02

    Prototype

    A working version, tested on realistic inputs, before any large commitment.

    • Built against the agreed success measures
    • Results shared openly, including what does not work
    • A clear decision point: continue, change course or stop
  3. 03

    Build

    Production engineering, done properly.

    • Integration with your systems and data
    • Security, access control and logging
    • Thorough testing before anything goes live
    • Documentation your team can use
  4. 04

    Operate

    Launch is the start of the work, not the end.

    • Monitoring of quality, speed and cost
    • Careful release of changes
    • Either ongoing operation by us, or a full handover

What you can expect from us.

Measure before claiming

We agree how output will be judged before building, and we report results against that, not against impressions.

Small, reversible steps

Work is delivered in small steps that are easy to check and easy to undo, so a problem stays small.

People stay in charge

Systems we build have clear limits, and sensitive actions can require a person to approve them.

Your data stays yours

Client data is used only for that client's project and is handled under the terms agreed in writing.

Plain communication

Regular written updates in plain language: what was done, what was learned and what comes next.

Software you can own

Readable code, written documentation and no unnecessary lock-in, so your team can take over at any point.

Ways to engage.

Every engagement is set out in a written agreement before work begins.

01

Fixed-scope build

A defined product or feature, delivered against an agreed scope and timeline.

02

Ongoing development

A continuing arrangement to build, improve and operate a system over time.

03

Review and advice

An independent look at an existing AI system: reliability, cost, quality and risk, with written recommendations.

We are not right for every project.

Better to know early. Here is roughly where we do our best work, and where we do not.

A good fit

  • You have a real problem, and people waiting on the answer.
  • Reliability matters more to you than novelty.
  • You want a partner who will say so when something is not going to work.
  • You would rather see a working version early than a long plan.

Probably not

  • You need a large team working on site.
  • You want a demo for a presentation and nothing after it.
  • The goal is to use AI for its own sake.
  • You need someone to start tomorrow with no scope at all.

Common questions

How does a project start?

With a message. Tell us what you want to build and we will reply by email, usually with a few questions. If it looks like a fit, we talk it through and send a short written proposal.

What size of project do you take on?

Anything from a focused prototype to a complete product. If something is too small or too large for us to do well, we will tell you.

Who owns what you build?

That is agreed in writing before work starts. For custom builds, our usual position is simple: the client owns what the client paid for.

Will you sign a confidentiality agreement?

Yes. We are happy to sign one before you share any detail.

Can you work alongside our own team?

Yes. We can build with an in-house team, or take a project on from start to finish and hand it over when it is running.

Do you build products of your own as well?

Yes. Running our own products is where much of our practical experience comes from, and it keeps us honest about what works.

Have something you want to build?

Tell us about the problem, the people who will use it and any constraints. We will reply by email.