Home / What we build

Software that earns its place.

We develop products end to end. Here is what that covers, what goes into each type of work, and what you should expect to receive.

  1. New products

    The starting point is a problem and a group of people who have it. The end point of the first phase is a release that lets those people complete one important task more easily than before. Between the two sits discovery, design, development and release, each described on its own page.

    Typical inputs: a description of the problem, access to intended users, any existing material and a decision maker. Typical outputs: a product brief, tested journey sketches, a working web application with accounts and the data model it needs, an admin view for your team, automated tests, and deployment to hosting registered to you.

    Limitations: a first release is deliberately narrow. If your plan relies on dozens of features arriving at once, we will challenge it.

  2. Internal tools

    Software for your own staff has a peculiar advantage: the users are down the corridor. That makes it possible to test ideas within days and to measure success in hours saved or mistakes avoided.

    Common examples include scheduling and allocation boards, multi-step approval flows, job and case tracking, stock or asset registers and simple customer records. Very often there is a spreadsheet that has been stretched far past its design, with formulas only one person understands. We map what the spreadsheet actually does, including the unwritten rules, before replacing it.

    You receive: a tool shaped around the real workflow, permissions that match who should see what, imports of the existing data where it is clean enough to move, and a short guide for staff.

  3. Prototypes that answer one question

    A prototype is an experiment with a hypothesis attached. Before making one we write down what it is meant to prove, who will see it and what result would change our minds. Desirability prototypes are clickable sketches; feasibility prototypes are rough technical tests against a real system or data source.

    We are careful to label prototypes clearly and not to let them drift into production. Code written to learn quickly is rarely code you want running your business.

  4. Technology choices

    We choose mainstream, well-supported tools that other developers can maintain, and we explain the choice in writing. Novelty is not a reason. Where your organisation already uses particular systems, fitting around them usually beats replacing them. We do not claim formal partnerships with any platform vendor.

"Could a person do this by hand for a month while we learn?" is our favourite question.

Plenty of features can start as a manual process behind a simple form. It is not glamorous, but it means your first release arrives sooner and you only automate what proves its worth. We will point out where this applies to your plans.

Slice your own first release

Not sure which type fits?

Describe the situation in plain words and we will suggest a starting shape.

Start a brief