Rebuilding a decade-old mortgage system so 4,000 advisors could trust their own tool again.

Through stakeholder interviews, contextual enquiry, and ongoing collaboration with subject-matter experts across UK branches, I built a detailed picture of who uses the mortgage system and why it was failing them. Three distinct user profiles emerged. My design work focused on the primary user — the Certified Banking Manager who processes mortgages daily and carries personal liability for every application.

After the discovery phase, I analysed every artefact — SME interviews, screen-share recordings, regulatory documents, workflow maps — and proposed high-level wireframes showing how the entire system could be restructured. Not a polished prototype. A vision. A way to say: this is what it could feel like.

That deliverable gave the project sponsor and senior stakeholders enough confidence to greenlight development. It also gave both teams — consultancy and bank — a shared language. Before the wireframes, conversations about the system were abstract. After, they were concrete. Everyone could point at the same thing.

From that point, I moved into deeper journey-level design. But my primary tool — Figma — was stuck in a 4-month UK compliance approval process. Rather than stall, I built prototypes that referenced the coded component library already available to developers internally. I mapped each flow to the Swedish-managed design system, ensuring developers could start building while I worked toward full Figma access. When it arrived four months later, I redesigned key flows using the approved components — but by then, development was already underway. That decision not to wait meant we didn’t lose a quarter to process.

This wasn’t a project where I received requirements and delivered wireframes. I was embedded in the delivery organisation — shaping how design, research, and product decisions were made across the entire programme.

Testing wasn’t confirmation. It was the engine of the design process. Every session surfaced something we hadn’t anticipated — and the design changed because of it.

And then we had to remove the feature they valued most.

A couple sits across from you. Second home. Complex income. Non-standard mortgage. You open the system. The form already knows this is a joint application — only the relevant sections appear. You capture the first applicant’s details. The header tells you whose data you’re entering. You scroll through the sections — everything is visible, nothing hidden behind tabs. You switch to the second applicant. A phone call interrupts you. When you return, every field is exactly where you left it. Nothing is lost. Before generating the compliance documents, you enter read-only mode and scan the whole application. Everything is correct. You generate, review, sign.

That’s the distance this project travelled.

Trust is the currency, not pixels. In a department that wasn’t used to working with a UX designer, my first job wasn’t to redesign screens — it was to prove that user-centred thinking produces better outcomes. I did that with one set of wireframes that got the entire programme greenlit. After that, the door was open.

The real experts sit in the branches. The SMEs who’d been processing mortgages for 15 years knew every edge case, every workaround, every fear. They didn’t just validate my designs — they shaped them. The best design decisions on this project came from listening, not from innovation workshops.

What I’d do differently: I’d push harder for early design system alignment. The evolving component library created rework that was avoidable with closer collaboration with the Swedish design system team from week one. I’d also establish a cross-team design review cadence sooner — I introduced it midway through the project, but the consistency benefits would have been greater from the start.Draft