Based in Sliema, Malta
MSD
Collapsing a 71-field exam request into three steps.
A clinical workflow where one late error cost the entire form. I rebuilt it on the front end without touching the backend or rewriting a line, and cut its modeled friction by 66%.
Summary
Context
PD Point is the clinical-protocol platform MSD's network runs an exam on end to end. It serves 9 profiles across 134 HTML pages, from request to material return. The brief asked to renovate the visual and the UX, fast, without touching the backend or migrating to React. The request screen was the worst of it. A wall of ~71 fields, where one error caught at the bottom cost the entire form.
What I did
Reframed the brief and sold the approach.
I recast the request from a reskin into a UX problem and brought stakeholders onto the new approach before any pixels. That bought the room to slice the flows instead of only re-dressing them.
Mapped and sliced the critical user journeys.
I designed the core journeys and broke them into steps only where it helped. A friction index built on Hick's and Miller's Law showed where slicing paid off and where it would slow heavy users, so the cuts landed where they did not hurt the flow.
Chose MUI as the field reference to buy time.
I picked Material UI for the fields on purpose. It is heavily documented and visually close to MSD's current CMS design system. That spared me from drawing forms from scratch and freed my time for the real work: analyzing and slicing the journeys.
Team
Client
MSD
Role
Product Designer
Technologies
Legacy system
Responsibilities
Research & Synthesis · UX / Content Strategy · System Design · Legacy Integration (HTML delivery) · Design System Documentation
Results
Risk cut, backend untouched
A front-end-only overhaul, validated on the core journeys before scaling.
−62%
decision density (~71 fields read as ~27 real decisions).
−66%
modeled friction on the request, same content.
Zero
backend changes. Shipped as a front-end layer.
The friction figures come from a modeled friction index (Hick's Law + Miller's Law) rather than a live usability test. Direct access to the live system and to real end users was blocked for legal reasons, so I validated the work against a domain-expert proxy and a pilot before scaling.
Shipped and scaled.
Mobile-ready foundation
The product got its first responsive, mobile experience, with a design system and documentation so improvements would not drift into isolated, inconsistent pages.
Sustainable handoff
Delivered as HTML and CSS with stable field IDs and backend-ready tables, so engineering could plug the legacy data straight in and scale pages without a redesign per adjustment.
The client came back
The pilot confirmed the direction, the redesign scaled platform-wide, and the client returned for a second project, the clearest endorsement of all.
Details available upon request. Contact me via LinkedIn.