Generative
design.

Identity systems built on rules, not prompts. One system makes thousands of on-brand layouts, patterns or images without a designer redrawing each.

A smartwatch on a wrist in daylight

What’s included

(05)

Your brand needs more versions of itself than any design team can draw by hand, and image generators drift off-brand.

  1. The rules of the system

    Colour, shape, type and spacing written as rules a program can follow, so every output is recognisably yours.

  2. Patterns and graphics

    Backgrounds, textures and shapes that are different every time and still belong to the same family.

  3. Layouts at volume

    Social posts, banners, packaging variants or covers produced from your content, in every size you need.

  4. Data-driven visuals

    Graphics that change with real inputs, such as a name, a place, a number or a date.

  5. A tool your team runs

    A simple interface so marketers can make new outputs themselves, inside the limits the system sets.

When you need it

  • You need hundreds of on-brand assets and the design queue cannot keep up.
  • Every market, product or customer needs its own version.
  • AI image tools give you things that look nothing like your brand.
  • You want an identity that moves and changes, not one fixed logo.

How it runs

(04)

You get a written scope, a fixed price and a start date before anyone begins.

  1. Define

    We take your identity apart and write down what must stay fixed and what may vary.

  2. Code

    The rules become a working generator, tested against your real content.

  3. Tune

    We look at many outputs side by side and tighten the rules until none look wrong.

  4. Equip

    Your team gets the tool, the source code and a guide to using it.

What you leave with

  • A written rule set for the identity
  • A working generator for patterns, layouts or images
  • A library of approved outputs to start with
  • A simple tool your team can use
  • Source code and files, handed over

Start your project with us

FAQ

Answered questions.

A brand identity written as rules that software follows. Instead of one fixed artwork, the system produces many variations that all look like the same brand. The designer sets the rules and the program does the drawing.

No. Image generators guess from a prompt and drift. A generative system follows rules you approved, so the results are consistent. Machine learning can be part of it, but only inside those rules.

Not always. We can build rules from your current identity. If the identity itself needs work, that starts with our brand practice.

Portrait in profile against a red ground