2026
Redesigning the credentialing experience inside a medical business-administration platform used by 1.4M+ providers.
Role
Offshore design track lead
Team
6 Designers, PM, PO & Leadership
Timeline
10 mo
Tools
Figma (Design), Maze (Usability Testing),
LLMs for research and processing
TL;DR: The Numbers
Setting the Stage
Millions of practitioners rely on it to:
Submit professional information
Maintain accurate records across hospitals
Stay compliant with regulatory requirements
The Constraints
Compliance is
non-negotiable
Past affiliations must remain auditable.
No deletion. No shortcuts.
Legacy data is real
Millions of records already exist,
Messy, inconsistent, but critical.
Problem
Doctors were re-entering the same information repeatedly
Historical records existed, but weren’t reusable
Errors were discovered late, during admin review
Admin teams spent time fixing data instead of reviewing intent
Zooming In
To show where this hurt most, here's the flow I went deep on
A "hospital affiliation" is just the relationship between a doctor and a hospital, past or present.
Each doctor lists these kinds of connections:
Hospitals where they can admit patients: places they actively work at, with full privileges.
Hospitals they're connected to but don't admit at: partial roles, research positions, consulting.
Old hospitals they used to work with: kept on file for legal and credentialing reasons.
If it's wrong, a doctor might not be legally allowed to practice where they say they are.
Who it's for
A doctor managing their own credentials, no training, no patience, expecting it to just work.
Admins and ops teams used the same system, but the doctor was the one abandoning the flow, so they're who I built around.
Legacy Experience
While legacy experience had been in existence since a long time it needed some remodeling and user flow arrangement.
Screens are anonymized and heavily blurred to comply with NDA restrictions.
All affiliation types were shown at once, increasing cognitive load before users understood the task.
Long, scroll-heavy layouts obscured task progress and completion.
Absence of progress indicators reduced confidence in whether steps were finished.
Information was organized by system structure, not user mental models.
Error states lacked visual hierarchy, making issues easy to miss.
How I Researched
We were stuck here:
Where in the flow do people
lose confidence,
make mistakes,
or give up?
How we approached it & Method
I ran a mixed-methods study on the legacy flow, combining quantitative usability metrics with qualitative behavioral observation to find both where users failed and why.
With a span of 70 participants across both personas (providers and credentialing admins), recruited to match real-world experience levels.
Mirroring real affiliation workflows: adding a new active affiliation, updating an existing record, and entering a past inactive affiliation.
After each scenario to capture perceived difficulty.
To benchmark the legacy experience against a comparable baseline.
What it tracked:
Quantitative:
Task success rate — completed without assistance, with assistance, or failed
Time-on-task per scenario
Error rate — misclicks, wrong-field entries, abandoned steps
Drop-off points — where in the flow participants gave up or got stuck
SEQ and SUS scores for benchmarking
Research Findings
Task Success Rate
Users abandoned
Needed assistance
Component Design
Previous data was a long scroll. Can we condense it into tables for better data representation?
Component Design
Taking it one task at a time and taking care of all use cases
Component Design
Lets avoid double data entry due to typos or change in the input style.
Final Screens
This is where practitioners start their flow
Final Screens
Optimising for scannability
Final Screens
Lets aim to save the data than to realise you messed up later
Each tab focused on a single task at a time, lowering cognitive load
Added a visible progress indicator, giving users a clear sense of completion
Reduced scrolling by moving content into tables and controlled views.
Improved error discoverability by surfacing issues
Impact
Early validation showed strong improvements in both usability and confidence.
participants said they’d prefer this flow over the legacy experience
30% to 9%
Drop-off fell
task success rate in completing hospital affiliation updates
Reflection
The system already had the data — the real challenge was helping users trust and reuse it.
Leadership meant owning clarity, not just screens.
As the flow grew in scope, my role shifted from execution to ensuring the experience stayed coherent across the team.
FAQ