UX Center of Excellence
Built a UX Center of Excellence from zero. It delivered 80% user adoption and cut accessibility issues by 70%, inside a WCAG/ADA-regulated healthcare environment.
build a scalable, accessible, and consistent design system from zero, in a regulated healthcare environment, without disrupting ongoing product development?
There was no design system, no component library, no naming conventions in Figma. Accessibility complaints were surfacing in production, and in healthcare, compliance failure has legal consequences, not just UX ones.
There were no documented design principles. PMs were making product decisions without research backing. The design function existed but lacked the infrastructure to scale or defend its decisions.
How the Component Library Felt to Use
The production member and admin portals are client-confidential. Below is an interactive recreation built for this portfolio, using the same interaction pattern with a generic product name and fictional data.
Illustrative recreation: interface rebuilt from memory in plain HTML/CSS/JS. Product name, brand names, and data are all fictional.
Every specimen shipped with a documented focus ring, a minimum 4.5:1 contrast ratio, and a written usage rule — the same three checks any component had to clear before it entered the library.
| Plan | Members | Status |
|---|---|---|
| Plan A — Medical | 412 | |
| Plan B — Dental | 288 | |
| Plan C — Vision | 150 |
Built a WCAG/ADA-aligned component library in Figma from zero, prioritizing accessibility-first components over visual refinement. I framed the design system as a risk-reduction tool to get cross-functional buy-in.
Established a design-to-dev handoff process and trained PMs and stakeholders on design system usage. Redesigned the shopping experience flow, which became the flagship proof of adoption: 80% of the product now runs through the system.
I prioritized accessibility-first components over visual refinement. In healthcare, compliance failure has legal consequences. Framing the design system as a risk-reduction tool, not a design preference, is what got buy-in from legal, engineering, and product at the same time.
Building from zero is different from inheriting a system. The hardest part isn't the design. It's getting cross-functional buy-in before there's anything visible to show. The first deliverable that mattered wasn't a component either. It was a shared vocabulary.