Product Designer: AI Powered Interfaces

Most AI products today do too much. They surface information users did not ask for, trigger actions without explanation, and replace familiar flows with unpredictable ones. The result is not intelligence. It is noise.

This role exists to fix that.

You will design interfaces where AI does real work in the background and the user stays in control of what happens. That means knowing when to show the AI and when to hide it. When to guide and when to get out of the way. The work spans research, interaction design, visual execution, and business judgement. You will need all four.

What You Will Work On:

• AI assisted flows: Design interfaces where machine learning outputs — recommendations, risk flags, automated decisions — are surfaced to users without creating confusion, anxiety, or loss of control

• Onboarding and activation: Own the flows that bring new users from zero to first meaningful action, with particular focus on clarity at moments of financial decision making

• Voice and conversational UX: Design interaction patterns for voice and chat based entry points, where the absence of a visual interface makes clarity of language and flow structure even more critical

• Accessible and inclusive design: Build for users across geographies, languages, and levels of financial literacy, accessibility is a structural requirement here not a final checklist item

• Design system ownership: Create and maintain components and patterns that scale across a global multi market product without fragmenting into inconsistency

• Research and experimentation: Run usability sessions, analyse funnel data, design A/B tests, and translate findings into concrete design decisions with measurable outcomes

5 Non-Negotiable Requirements

1. AI interface design in production
You have designed interfaces that surface machine learning outputs to real users: recommendations, scores, flags, or automated decisions. You made deliberate choices about when to show the AI reasoning and when to abstract it, and you can explain why.

2. End to end ownership of complex flows
You have owned a multi step product flow from research brief through to shipped and measured outcome. Not contributed to one. Not handed off after wireframes. Owned it.

3. Design decisions grounded in data
You have used quantitative signals — funnel drop off, conversion rates, error rates, time to complete — alongside qualitative research to make and prioritise design decisions. You can show work where data changed what you built.

4. Fintech or regulated product experience
You have designed for financial, payments, insurance, or similarly regulated products where the cost of confusion is not just frustration but money lost or trust broken. You understand what KYC, AML, FX, and fee transparency mean for a user flow.

5. Accessible and inclusive design practice
You have designed for users across different contexts, literacy levels, languages, or abilities as a core part of your process, not as a post launch audit. You can articulate specific decisions you made and why.

What We Are Looking For Beyond the Requirements

Calm design instinct: You default to reducing cognitive load, not adding features. When in doubt you remove, not add.

Ethical interface judgement: You recognise dark patterns when business goals push toward them and you propose alternatives rather than just flagging the problem.

Cross functional fluency: You work directly with engineers, product managers, compliance, content, and legal. You communicate design decisions without jargon and without defensiveness.

Voice and conversational design experience: Familiarity with designing for chat, voice, or LLM powered interfaces is a strong advantage.

Figma mastery: Component architecture, auto layout, variables, and design token management at a level that makes handoff frictionless for engineers.

What This Role Is Not

Not a visual design role. Aesthetic execution matters but it follows research and interaction decisions, it does not lead them.

Not a generalist UX role. We need someone who has worked in high stakes, regulated, data heavy product environments.

Not an agency role. You will own outcomes, not deliverables. Shipped screens are not success. Behaviour change is.

How to Apply

This role is open to EU based candidates only. We are not considering applications from outside the European Union at this time, regardless of remote working arrangements or timezone compatibility.

Submit your CV, a link to your portfolio, and a short video answer to one question:

Walk us through one AI powered or data driven interface you designed. What did you decide to show the user, what did you decide to hide, and why?

A portfolio without at least one end to end case study will not be assessed. A case study without your specific contribution, the trade offs you navigated, and the measured outcome will score the same as no case study.

The video format is uncomfortable. We know that. If you still do it, that already tells us something.

How Applications Are Assessed

I want to be upfront about how this works before you invest your time.

Every application is scored against the 5 non-negotiable requirements only. One point per requirement. 5 out of 5 to proceed. Not 4. If a requirement appears in your CV or portfolio without specific context describing what you owned and what it affected, it scores 0.

I compare all applications before advancing anyone. If more than 15 candidates score 5 out of 5, I rank by depth of AI interface experience, scale of product owned, and quality of case study documentation. The top 15 go forward. If fewer than 15 score 5 out of 5, all of them go forward.

The video is reviewed by me and the team together. We are not assessing camera confidence. We are assessing specificity, personal ownership, and design judgement under a real constraint.

I do not follow up to ask for clarification on an ambiguous application. What is submitted is what is scored.

You will hear back from us regardless of outcome. That is a promise, not a pleasantry.

Hubert Warszta
Tech Recruiter | WhyHireWrong? | SIGL:IO