Skip to content
Work with us
All services

Clickable, interactive designs at an early stage that are nearly indistinguishable from the final product.

Duration: Typically 5–15 days

A prototype realistic enough that people react to the product rather than to the placeholder. We make clickable prototypes with real content, working states and the micro-interactions that make an interface feel like software instead of a slideshow. We use them for user testing, for stakeholder demos or to settle a question before development starts.

  • Clickable prototype
  • Micro-interactions
  • Test scenarios
  • A settled question

AI makes the prototype in little time, so people react to the product instead of a placeholder; we keep it true to the design and decide what it has to answer.

  1. Scope

    We agree on the question the prototype has to answer and the tasks a tester goes through to answer it. Those tasks work from start to finish; anything outside them is on screen but does not respond.

    What we do
    Set the question the prototype has to answer.
    How AI helps us
    Lists the tasks the prototype needs.

  2. Make

    Clickable, with real content and working states rather than lorem ipsum.

    What we do
    Keep the prototype true to the design.
    How AI helps us
    Makes the prototype in little time, with real content.

  3. Detail

    We add the details whose absence gives a click-through away: a button that shows it is working, a confirmation after saving, a screen that slides in instead of jumping.

    What we do
    Judge which details make the prototype feel real.
    How AI helps us
    Adds the transitions and feedback states.

  4. Use

    We put it in front of users in a test, or stakeholders in a demo. Where people get stuck or hesitate, we change the design, and the next round tests the new version.

    What we do
    Change the design where people get stuck.
    How AI helps us
    Clusters what testers and stakeholders say.
Work with usMore on AI

  • More and more often in code. With AI we can make a prototype in little time that already comes close to the real product: real screens, real interactions and, where it matters, real data. When the question is about a single flow and Figma is faster, we use Figma.

  • Much of it, yes. The heavy lifting is in the flows and interactions, and by the end of the prototype those are thought through and tested. We usually work from a design system and keep the prototype as close to the real product as we can, so screens and components often carry straight over. Expect one more round on top: a prototype is there to gather feedback, and feedback changes things.

  • Real enough that a tester does not notice it is a prototype. That decides what you get back. If it looks fake, the feedback is about the fake: 'we don't use photos like that', 'this text means nothing to me' or people who cannot picture themselves in a gray sketch. If it looks real, they react to the product itself: whether they find what they are looking for, and whether they understand what is happening.

Is your question not here?Jasper is happy to answer it.

Ask a question