01
Understand
Evidence before opinion.
Usability testing · study setup (prototypes, tasks, questions) · eye-tracking · heuristic evaluation · accessibility (WCAG 2.1) · stakeholder and business requirements.
Portfolio — 2026
Scroll to start the journey
The page builds as you scroll
I make complex journeys feel simple.
My job at BMW was to make buying a car online feel easy: no distractions, no doubt, just the next clear step. This site is the other half of me, where flair meets focus. Same designer, two modes.
Human by name. Human-centred by design.
What I bring
01
Evidence before opinion.
Usability testing · study setup (prototypes, tasks, questions) · eye-tracking · heuristic evaluation · accessibility (WCAG 2.1) · stakeholder and business requirements.
02
From first flow to production.
User flows · wireframes · prototypes · high-fidelity UI · design systems · stories and acceptance criteria · refinement with Product Owners, Business Analysts and developers · visual QA · environment testing.
03
Ideas at production quality, fast.
AI image and video direction (Nano Banana, Seedream, Seedance 2.0 / Higgsfield) · identity and product consistency with reference workflows · prompt libraries · motion and scroll storytelling · AI-assisted build with Claude Code.
About · Story
I started in Interior Architecture, designing how people move through space. Today I design how they move through screens. The principles never changed: hierarchy, flow, and putting people at the centre of every decision.
Working in long-running product teams taught me that good UX goes far beyond the design file. I've been part of every stage, from the first concept to the production release, and I've learned to love both ends: the rigour of research, and the craft of making something people want to look at.
Design & product
AI & creative

Selected work · Focus
Screen archive
The live checkout, step by step, per brand and on both desktop and mobile. Open a screen to scroll through it, then use Next to walk the whole journey.
Select an archive
Click a screen to open it
Evolution
The same checkout steps, from the version that was live when I joined to the one live today.
Testimonials

Contact
Product design, research, design systems or AI-assisted creative: I'm always happy to talk.
quewinhuman@gmail.com (opens your email app)
Freelance & studio work under QH Dimensions
01 · UX case study
01
Buying a vehicle involves much more than selecting a car and paying for it. DCS needed to bring the different parts of the purchase process together into one guided experience, while adapting to different customers, vehicles and sales journeys.
What the customer asks
What the checkout has to handle
Behind that seemingly simple journey was a much more complex product ecosystem:
02
DCS wasn’t a single checkout page. It was a living product spanning multiple journeys, brands, markets and feature teams. As one of a small number of designers, I worked across the end-to-end journey, helping keep the experience consistent as the product evolved.
How my role grew
Early days
Joined DCS in 2021 and quickly became involved across multiple parts of the product. The team was small, which meant designers contributed well beyond individual screens.
Scaling the product
As the platform expanded, work became organised across multiple feature teams covering different parts of the checkout ecosystem. I worked across both DCS and Trade-In, contributing to a wide range of journeys and features.
Today
My focus shifted fully to DCS, while continuing to work across multiple feature teams. Rather than owning a single feature, I helped maintain consistency across the complete checkout experience and improve the user experience through user research and testing.
How a feature moved through DCS
Select a phase to see what happened and what I did.
01 / 06
What happened
A new feature or improvement was identified through business needs, customer feedback, usability findings or market requirements.
My role
I explored possible solutions, mapped user journeys, created flows and early concepts, and aligned on the problem we were solving before moving into design.
02 / 06
What happened
The initial concept was refined with Product Owners and Business Analysts to align customer needs, business requirements and technical constraints.
My role
I worked closely with POs and BAs to clarify requirements, refine user flows and ensure the customer experience remained coherent as requirements evolved.
03 / 06
What happened
The feature was reviewed and refined with developers, testers and the wider feature team before sprints.
My role
I presented designs, answered questions, refined edge cases and prepared implementation-ready designs and specifications.
04 / 06
What happened
Development began and designs were translated into production-ready functionality.
My role
I supported implementation, reviewed builds, answered development questions and worked closely with engineers to ensure the intended experience was correctly realised.
05 / 06
What happened
The feature was tested before release to identify issues and ensure production readiness.
My role
I conducted visual QA and took part in testing on every environment before the release was pushed to production.
06 / 06
What happened
The feature was released to customers and became part of the live checkout journey.
My role
I took part in testing in production, gathered feedback, supported follow-up improvements and fed learnings back into future iterations.
One journey, many teams
DCS was built through autonomous feature teams, but the customer experienced it as a single journey. A key part of my role was helping maintain consistency across the checkout as different teams contributed to different areas.
The teams I designed for, and the badges I designed for them.




03
I contributed to and helped maintain the DCS component library as it evolved across multiple brands and journeys, before supporting the transition from the custom DCS library to BMW’s NEO design system. The experience stayed consistent while each brand retained its own identity.
One checkout, three brands
Same patterns, different identities
The underlying components and interaction patterns remained consistent across brands, while visual design, tone and styling adapted to each brand’s identity.
Drag the divider to compare BMW and MINI on the same step.
This shared foundation is what made the product evolution in the next chapter possible.
04
DCS was never redesigned in one go. Release by release, it grew from a long accordion checkout into a premium but content-heavy experience, and finally into a focused, guided, multi-step journey. Most of the decisions behind that meet on one page: the welcome step.
Explore the evolution
Select a numbered hotspot to explore a design area, then use the slider to follow how it evolved across versions.
View details about each change below the page viewer.
Earlier version
The vehicle image sat inside the summary cart on the right, together with pricing, vehicle details and specifications.
Progressed version
The vehicle moved into a large hero area at the top of the page: bigger, and on a visual backdrop. The goal was to make the checkout feel more premium and emotionally engaging by keeping the customer’s chosen vehicle front and centre.
Current version
The vehicle moved back into the summary on the right, but kept the premium feel: visual backdrops, larger imagery, an image gallery and zoom.
Why it evolved to what’s live today
The drive to create a more premium online car-buying experience led us from the Earlier version to the Progressed version. While the larger hero area looked impressive, it pushed key checkout content further down the page. Customers had to scroll every time they moved between steps, making it harder to stay oriented and slowing progression through the journey.
Keeping the user’s scroll position between steps would have meant the hero was rarely visible, undermining its purpose. Through iteration, we realised the original structure wasn’t fundamentally broken. What it needed was refinement rather than reinvention.
What it solved
Earlier version
An accordion: every step lived on a single page and opened and closed in place.
Progressed version
The accordion got a stronger hierarchy: a more prominent active state, disabled future steps and a clear difference between active, completed and upcoming sections. It was the first stage of stronger progression cues.
Current version
The accordion became a horizontal stepper at the top of every page.
Why it evolved to what’s live today
The accordion worked well initially, but became harder to scale as more information and journey complexity were introduced. We explored ways to improve the accordion itself, but ultimately found that a stepper offered a clearer mental model.
A stepper keeps progress visible, makes navigation easier to understand, and gives customers a stronger sense of progression throughout the checkout journey. It also frees up page space for the content that matters most.
What it solved
Earlier version
One long page: the customer saw every step at once. As content and functionality were added, the page grew with them, especially on Personal data, Finance, Order overview and Order confirmation, and even more on mobile.
Progressed version
The accordion grew further, and Trade-In became the first step as the feature grew in importance.
Current version
The checkout became page-based: each page focuses on a single task.
Why it evolved to what’s live today
Focus. Showing only the current step reduced visual noise and helped customers concentrate on the task in front of them rather than the entire journey.
What it solved
Earlier version
Login lived inside the Personal data section. Customers could land on a step called Personal Data while being asked to log in, which caused confusion.
Progressed version
Still inside Personal data, but with more ways in: create a BMW ID, log in or continue as a guest. That reduced friction and let customers explore the checkout before committing more strongly later in the journey.
Current version
Pre-login became a dedicated welcome page before checkout begins. The stepper previews the journey without activating a step; Personal data only becomes the first active step after login.
Why it evolved to what’s live today
Onboarding and checkout serve different purposes. Separating them removed the confusion of logging in on a step called Personal Data, and gave the checkout a proper start.
What it solved
Earlier version
There was no benefits section on this page yet.
Progressed version
The benefits of buying online were shown throughout the checkout, to keep reinforcing the value of completing the purchase online.
Current version
The benefits moved to the welcome page, before checkout starts.
Why it evolved to what’s live today
Customers only need the value proposition once. Moving benefits to the start of the journey kept the rest of the checkout focused on completing the purchase.
What it solved
05
Personal Data became one of the most complex parts of DCS. What started as a simple form evolved into a system that needed to support saved data, missing data, multiple addresses, multiple contact methods and a wide range of customer journeys.
What the page had to handle
Three versions of the same page
Choose a version, then a scenario, to see how the page handled it.
The original Personal Data experience lived inside the accordion checkout.
What we observed
As more scenarios were added, complexity accumulated inside a single page.
Personal Data became a page of its own behind a stepper, and the interface got lighter.
What improved
What hadn’t changed yet
We simplified the UI, but we hadn’t simplified the problem yet.
Personal Data starts as a summary, and every bigger task gets a focused page of its own.
What changed
Customers now focus on one task at a time instead of managing all personal data at once.
The constraint that shaped the solution
A BMW ID can be created with just a name, a surname and an email address. So customers can arrive in DCS with important information missing: no valid address, no valid contact number, or incomplete personal details.
That’s why not every interaction could move into a separate, focused flow. When mandatory data is missing, its section opens directly on the Personal Data page, so the customer can complete it before moving on. Addresses are also checked as soon as the customer enters DCS, not only on Continue, so an invalid address opens straight away.
Personal Data wasn’t redesigned because the UI looked old. It evolved because every release introduced new customer scenarios, business requirements, accessibility improvements and usability learnings.
The final experience is the result of years of simplifying complexity rather than adding features.
06
Personal Data was only one part of the challenge. Across the wider checkout journey, customers needed confidence in their decisions before they could continue.
08
DCS wasn’t designed for a single screen size. Every journey had to work across mobile, tablet and desktop while supporting multiple markets, translations and branding requirements. The challenge wasn’t creating separate experiences, but scaling one consistent checkout system across many contexts.
Long translations, varying content lengths and design-system constraints regularly challenged layouts. Working within BMW’s NEO design system meant solutions needed to remain flexible without relying on custom components for every edge case. The result was a checkout designed to adapt gracefully across devices, markets and content variations.
Resize the checkout*
Drag the handle on the right of the window, or pick a device, and watch the checkout adapt.
Switch the brand to see the same components in another identity.
360 px · Phone
*For illustration only: a recreation built for this portfolio from the original designs. It uses stand-in typefaces and simplified components, not the official BMW, MINI or BMW Motorrad design system.
What influenced the design
Whether viewed on a phone, tablet or desktop, the goal remained the same: helping customers complete a complex vehicle purchase with confidence.
09
Impact
100,000+completed checkouts“Quewin distinguished himself through a consistent, rigorous focus on the end user.”
DCS didn’t evolve through one big redesign.
It evolved one problem, one insight and one iteration at a time.
02 · UX case study
01
For many buyers, a new car depends on what happens to the old one. Trade-In had to give a trustworthy, non-binding estimate in a few steps, without getting in the way of the purchase.
What the customer asks
What makes it hard
The value depends on the exact car: its model, engine and equipment. But identifying a car takes data that customers don’t always have, that differs from market to market, and that doesn’t always produce a match.
02
Trade-In grew out of DCS, and I worked across both products as part of the design team. My work spanned the Trade-In journey, and I collaborated across feature teams from discovery and exploration through refinement, implementation and delivery.
03
Trade-In didn’t start big. Its first version was a light section of the checkout, right after Personal Data. As trading in became more important, the step grew, and its place in the checkout moved with it.
Four versions of the step
Choose a version, then a screen, to see how the step grew and where it sat in the checkout.
The first version was deliberately small: one section of the one-page checkout, with a single way to identify the car. Customers landed straight in its fields and picked their car from a list, the manual search that later became one of five routes.
What it did
As Trade-In became more important, the same step became a full journey, with a start page of its own and, where the dealer took part, a non-binding estimate.
Why the start page came first
Empty fields for a step many customers skip only add to their cognitive load. A start page lets them decide first.
What changed
Then came the options. Three more ways to identify the car joined the manual search, and later a fifth, with each market choosing which to offer. With them, Trade-In moved to the very first step of the checkout.
Why it moved
Most of the new routes needed a logged-in customer, but the checkout still let guests buy. With Trade-In first, customers logged in before anything else.
Moving Trade-In first ensured the required customer data was available, but it introduced account access too early in the purchase journey, for a step customers might skip.
Asking for an account that early could make customers drop off. The checkout now opens with a welcome page of its own, and Trade-In moved back to the second step, after Personal Data.
What it looks like now
04
Not everyone knows their VIN. So identifying the car became a set of routes, from the most precise to a safety net, and every route ends in the same place: a summary to confirm the car, then the estimate. Each market switches on the routes its data supports.
Explore the routes
Select a route to see its screen, when it helps and what happens when something goes wrong.
The most precise route: the VIN helps identify the specific vehicle, including its model, engine and equipment.
When it helps
When the registration documents are at hand.
When it goes wrong
Few people know their VIN by heart. A help popover shows where to find it, the field checks for all 17 characters before searching, and when no car matches, a clear message says so.
A familiar alternative to the VIN, making vehicle identification easier in markets where licence-plate lookup is available. It follows the same pattern as the VIN route, so switching between the two requires little relearning.
When it helps
Available in markets with licence-plate lookup, with the exact lookup behaviour varying by country.
When it goes wrong
One plate can match several models. Instead of guessing, the customer picks theirs from a short list. An invalid plate, or no match at all, gets its own message.
For returning customers, an existing vehicle can be reused from MyBMW rather than identified from scratch.
When it helps
Returning BMW customers who are logged in. It’s the opposite of the manual search: it helps most when BMW already knows the car.
When it goes wrong
No car stored? The option switches off instead of leading nowhere. Several cars? The customer picks one. A first registration outside the car’s production period is flagged before it can skew the estimate.
No VIN or plate to hand: customers describe the car step by step, from brand and model to body, fuel and power.
When it helps
When there’s nothing to look up, or a lookup didn’t find the car. It’s the opposite of MyBMW: it helps when BMW knows very little about the car.
When it goes wrong
If the list of brands can’t load, a fallback keeps the search going. If no car matches the choices, the customer is told, rather than left with an empty list.
The safety net: customers type what they know, and the request goes to the dealer.
When it helps
In a market without any lookup, it’s the only route. Everywhere else, it catches the cars the other routes can’t find.
What happens next
There’s no automatic estimate here. The result says the value will be determined by the dealer, and answers the questions that raises.
Every route ends here: one checkpoint to confirm it’s the right car before it’s valued.
Why it’s there
A wrong match means a wrong estimate. The brand, model, fuel, power, registration and mileage appear together before the customer confirms.
Whichever route led here, the outcome is a non-binding estimate of what the car is worth.
What it explains
That the value is an estimate, not an offer; the car it’s based on; and what happens next. When no estimate is possible, the page says so and the dealer takes over.
A route that fails shouldn’t end the journey.
Every unhappy path leads somewhere: another route, a choice between matches, or the dealer.
05
As Trade-In grew, the valuation journey no longer needed to depend on a vehicle purchase. The same underlying experience was adapted into a standalone product, giving customers a way to value their current car before choosing their next one.
Inside DCS, Trade-In supports a vehicle purchase. Standing alone, the valuation becomes the destination itself: “What’s my car worth?”
One journey, two contexts
Switch between the checkout and the standalone product, on the start page or the estimate.
The challenge wasn’t rebuilding Trade-In.
It was making the existing journey meaningful without the checkout around it.
06
Alongside the core valuation experience, the team explored new product opportunities and smaller interaction improvements that could make Trade-In more useful and resilient.
01 · Eco-Bonus
During an in-person ideation workshop in Austria, we explored ways Trade-In could create value beyond the valuation itself. One concept became the Eco-Bonus: rewarding customers for trading towards a more efficient vehicle.
The concept connected the idea across the journey, introducing the benefit upfront and bringing it back alongside the final valuation with a more efficient vehicle suggestion.
01 · Introduce the idea
On the start page, before any vehicle details: what trading in can bring.
02 · Close the loop
Next to the estimate: a comparable, more efficient BMW and the bonus to earn.
02 · Skeleton loaders
Skeleton loaders reserve the structure of each page while data is loading, reducing layout shift and preventing controls from moving just as customers try to interact with them.
07
Trade-In grew from a lightweight checkout step into a reusable valuation journey: integrated within DCS when customers are buying a vehicle, and available independently when the valuation comes first.
How Trade-In evolved
Inside the checkout
From a lightweight optional step to a complete valuation journey within the vehicle purchase.
Many ways in
Multiple identification routes accommodate different customers, markets and available vehicle data.
Standing alone
The same valuation capability became a product of its own, independent of checkout.
Trade-In started as one step inside the checkout.
Designing for different customers, markets and imperfect data turned it into a journey that could stand on its own.