How I Work

From the first question to a build-ready front end.

Most designers tell you they have a process. I show you mine — recorded, stage by stage, on real projects. Here's the whole journey, from the first question to a build-ready front end.

✦ AI accelerates every stage

AI accelerates every stage. I lean on Claude and Gemini throughout — Gemini's large context window is a genuine unlock for big-data and deep-research work — to move faster through research and analysis. It doesn't replace the thinking; it speeds it up.

Understand the requirements

I start with the real world the product lives in: market research, comparative analysis, and the users' actual day-to-day tasks and usage patterns — including the existing systems they already rely on.

Market researchComparative analysisDay-to-day tasks & usage patternsExisting systems they rely on

✦ AI speeds up comparative research and market scans, so the landscape is mapped in hours, not days.

Map the landscape

I extract the core requirements and map or design the system architecture, filling gaps through interviews with stakeholders, the product owner, and the original developers.

Extract core requirementsMap or design system architectureStakeholder & product-owner interviewsOriginal-developer interviews

✦ AI helps structure requirements and reconcile what different stakeholders describe into one coherent model.

Gap analysis

I measure what exists against modern implementations, alternative solutions, and competitor products — and study how their users actually engage.

Against modern implementationsAgainst alternative solutionsAgainst competitor productsHow their users engage

✦ Large context windows let me read big comparison sets and engagement data in one pass — Gemini shines here.

Profile the users

I build personas grounded in demographics, software fluency, and real usage patterns — the actual humans, not a generic user.

DemographicsSoftware fluencyUsage patternsConsequential personas

✦ AI helps synthesise research into sharp, distinct personas that actually change design decisions.

Information architecture

I define screen flows, user journeys, and a framework overview — what can and can't be built, and where the boundaries can be pushed.

Screen flowsUser journeysFramework overviewBoundaries & where to push them

✦ AI helps pressure-test flows and surface edge cases before they reach a screen.

Wireframes & approval

I wireframe the core journeys to the business requirements, get high-level sign-off, and iterate.

Core journeys wireframedTo the business requirementsHigh-level sign-offIterate

✦ AI accelerates variations, so more options can be weighed before committing.

Design system & final UI

Corporate identity comes in; I build a reusable component library and design the full journeys to final UI, then get final approval.

Corporate identity inReusable component libraryFull journeys, final UIFinal approval

✦ AI assists with systematic, consistent components and token wiring across the library.

Build-ready front end

I produce the front-end UI in the developers' framework, looking and behaving exactly as intended — ready to hand off.

Front-end UI in the dev frameworkLooks exactly as intendedBehaves exactly as intendedReady for developers

✦ AI-assisted design-to-code shortens the distance between the final UI and working front-end.

What you walk away with

The deliverables I'm known for.

Design systems

Interactive prototypes (Figma + JS)

Specs & hand-off documentation

Structured research documents

See it on real projects