Skip to content

Design

AI interviews for product, UX and visual designers

Talk through your real design work with an AI interviewer, then see your results first and choose who sees them. Below: what strong answers show, the 5 practice questions, and a worked example.

  • Free for experts
  • About 15 minutes, whenever you like
  • Type or speak. No camera needed

Retake it up to 3 times: you keep every attempt, and choose which one represents you.

Who it's for

The work it covers

Product, UX, UI, visual and brand design, and design research.

Example roles

  • Product designer
  • UX designer
  • UI designer
  • Visual designer
  • Brand designer
  • UX researcher
  • Design systems designer

Name your exact specialty when you apply, and the questions follow it.

Topics the interviewer draws on in design

  • A design they shipped and what changed for users
  • User research that changed their mind
  • Exploring several options for a design problem and why they chose one
  • Critique or feedback that was hard to hear
  • Balancing user needs against business or technical limits
  • Accessibility and inclusive design decisions
  • Building or working within a design system
  • Handing a design to engineering and keeping the details right in the build

Scoring

What a strong answer shows

These are the words the scoring AI is given for design, as written in Aethel's configuration. The scoring AI is told to score an area below 60 when its answers show one of the warning signs, however confident the answer sounds.

Strong work

The candidate grounds decisions in research or testing with real users, handles accessibility on purpose, explains the options and trade-offs, and shows what changed for users with evidence.

Warning signs

The candidate justifies choices only by taste, trends or a manager's request, never involves users, treats accessibility as someone else's job, or cannot say whether the design worked.

The four scored areas

  • Design problem solving

    Problems, options, measured results

    Counts as evidence when the candidate describes a specific design problem, what caused it, the options they explored, a trade-off, or how they measured whether the design worked.

  • Craft and design decisions

    Flows, hierarchy, accessibility, research

    Counts as evidence when the candidate describes concrete design decisions: user flows, layout, hierarchy, interaction, accessibility, visual system or research methods, and why they chose them.

  • Communication

    Critiques, reviews, presenting work

    Counts as evidence when the candidate describes presenting or explaining design work to other people: critiques, design reviews, specs, prototypes, or walking stakeholders through a decision. Answers that respond to the question asked, in a clear order, with a concrete example, also count, but only up to 80 on their own. Grammar, word choice and fluency are never judged.

  • Collaboration

    Engineers, product, users, feedback

    Counts as evidence when the candidate describes working with other people: engineers, product managers, researchers, users or stakeholders, including handling feedback and disagreement. Saying "we" is not, by itself, collaboration evidence.

An area is scored only when the AI can quote you; otherwise it shows Not enough evidence, not a low score. How AI interviews are scored

Practice

The 5 practice questions

Questions like these come up in the interview. The free practice room asks 4 of them, at any level. The real interview follows up on your own answers instead.

  1. Question 1

    Tell me about a design you shipped, the problem it solved, and what you personally decided.

  2. Question 2

    What options did you consider and reject, and why?

  3. Question 3

    Imagine usability testing shows people can't find the main action on your screen. What do you change first?

  4. Question 4

    How would you defend a design decision to a stakeholder who strongly prefers a different direction?

  5. Question 5

    What would you improve if you had one more week, and how would you know it worked?

Answer 4 of them in the practice room

No account. Each answer gets a quick rule check in your browser.

Worked example

One answer, read by the rules

Made up by Aethel to show how the published rules read one answer: not a real candidate, and not scored by the AI. A real score can differ.

The practice question

What options did you consider and reject, and why?

The made-up answer

On a bike-share app in a mid-sized city, about 90,000 riders a month, the biggest support problem was people not finding how to end a ride, so they were charged for extra time. I owned the redesign of the in-ride screen and tried three directions. A floating button over the map was quick to build, but it covered the dock markers people were navigating to, and it failed contrast over light map tiles, so I dropped it. Ending rides automatically when the lock clicked looked ideal, but the lock sensor misread about one dock in fifty, and people would have been charged with nothing on screen telling them why. I chose a full-width status bar pinned to the bottom, with the ride timer, a large End ride button and a plain docked or not docked state, labeled for screen readers. In six moderated sessions, five people ended a ride in under ten seconds, against two of six on the old screen. Over the next three months, tickets about ending rides fell from about 400 a month to about 90, and the scooter team reused the same bar.

Marked words are the quotes the rules would look at.

How the rules read it

  1. The situationLevel 2 of 2

    A specific case with where it was and how big: one screen in a city bike-share app with about 90,000 riders a month.

  2. The reasoningLevel 3 of 3

    Two or more options compared, with why each was rejected: the floating button failed on the map and on contrast, and automatic ending failed on an unreliable sensor.

  3. The resultLevel 3 of 3

    Measured, and then followed up over time and adopted beyond the case: three months of ticket counts, and a second team reused the design.

Together the three ratings give 8 of 8 points, which is 95 on the fixed scale. Its longest complete quote has 51 words, so the quote rule does not lower it. It is evidence mainly for Craft and design decisions.

What would raise it

This already earns 8 points. A follow-up would probe what happened to the extra charges themselves, not only the tickets, and how the screen works for people who cannot see the timer.

The same story, told vaguely

I redesigned a screen in a mobile app because people were having trouble with one of the main actions. I explored a few different directions, tested them with users, and picked the one that worked best and was accessible. After launch, users had a much easier time and support issues went down.

A case without its context: no product, no audience size and no test numbers, so situation is 1 and the area is at most 60. 'Picked the one that worked best' gives no reason, and 'went down' has no number.

A whole result: the sample result. More examples: strong interview answers.

Company interview links

Roles companies can set up

Starter templates a company can use for an interview link, and what each one steers the interviewer toward.

  • Content designer

    Focus: Interface text, error messages and onboarding, voice and tone, testing wording with users

  • Design systems designer

    Focus: Component libraries, design tokens, documentation, accessibility, adoption across product teams

  • Product designer

    Focus: User research, interaction design, prototyping, working closely with engineers

  • UX researcher

    Focus: Planning studies, interviewing users, turning findings into product decisions

FAQ

Questions people ask

How do I talk about visual and interaction craft without showing it?

Describe the decision and the reason: what you changed in the hierarchy, layout, type or interaction, what it replaced, and what you saw in testing. Craft and design decisions are scored on concrete choices and why you made them. 'I made the button bigger' is thin. 'A full-width bar replaced a floating button that covered the map' gives the scorer something to quote.

What if my design decision was overruled?

That can be a strong answer. Say what you proposed, the evidence behind it, how you argued for it, and what happened. Collaboration covers handling feedback and disagreement with engineers, product and stakeholders. The field's warning signs include justifying a choice only by taste or a manager's request, so the evidence you brought matters more than who won.

I have no analytics for my designs. What can I use as a result?

Usability findings, task success in moderated sessions, support tickets before and after, or a researcher or engineer confirming what changed. The result rating counts a result confirmed by someone else, not only analytics. If all you have is an impression, say so honestly. It is read as approximate rather than measured.

Can I show my portfolio during the interview?

Not during the interview itself. It is a conversation, so you describe your work in words, and scores come only from what you say. You can add a portfolio link to your profile when you apply.

Do I have to share confidential details?

No. The interviewer is told not to ask for names or confidential details about clients, patients, students or employers, and to let you describe situations without identifying anyone. Describe unreleased work in general terms.

How to prepare for an AI interview

What this interview cannot do

It is one conversation of about 15 minutes, so it covers only part of what you can do, and we have not tested whether results predict job performance. The AI can misread an answer, which is why every score shows its quotes and people make the decisions. It is not a certification or license check. No independent bias audit has been done. Design was one of the 8 fields in our own test (not an audit) of whether non-native English grammar changes scores.

All known limitsYour rights as a candidate