From Manual Chaos
to Structured Enrollment
Redesigning healthcare workflows where every step matters.
Case Study
Lead Product Designer · 2024-2025
0%
faster enrollment
The Problem: Manual Everything
Healthcare enrollment was broken. Providers were spending hours on paperwork instead of patient care. Eligibility checks happened manually. Claims got rejected at high rates. There was no tracking, no automation, no visibility.
The problem wasn't policy. It was process.
I started with 30+ user interviews across providers, administrators, and clinicians. They all said the same thing: the system exists, but nobody has software to manage it.
Research: Understanding Real Workflows
User Interviews
I spent weeks talking to the people actually doing this work. Not stakeholders. Not executives. The staff manually processing enrollments, tracking eligibility, handling claims.
Key findings:
- ✓ Enrollment process took 8 weeks on average (mostly waiting)
- ✓ Staff didn't know eligibility rules — they guessed
- ✓ Service tracking happened in spreadsheets
- ✓ Claims were manually generated and submitted
- ✓ No feedback loop on rejections
Validation Rate
After identifying core problems, I proposed solutions. Users validated them early and often.
12
Proposed Solutions
Weekly
User Testing Rounds
85%+
Validation Success
The Solution: Structured Workflow
I designed an enrollment platform that turns manual chaos into structured workflow. No guessing. No spreadsheets. Just step-by-step clarity.
First name
Last name
Patient ID
Date of birth
Key Design Decisions
Guided Intake
Step-by-step forms that ask the right questions in the right order. No confusion.
Real-time Eligibility
Instant feedback on patient eligibility. Staff know immediately if the patient qualifies.
Service Tracking
Built-in capacity management for services. No more spreadsheets tracking what services are available.
Automated Claims
Claims generated automatically from enrollment data. Submitted with 100% accuracy.
How I Worked
Validating with users on real prototypes. Designing and coding in parallel. Just:
Understand
Interview providers, document real workflows
Prototype
Design interactive prototypes, test with users
Code
Build it directly into the codebase with AI assistance.
Validate
Real feedback from real users on real product
Refine
Iterate based on metrics and user feedback
Ship
Deploy to production, measure impact
Claims Automation
Claim forms were manually entered and prone to errors. No feedback. High rejection rates.
The Solution
I designed an automated claims engine that generates claims directly from enrollment data, validates them against rules, and provides real-time feedback on submission status.
Click a claim to see validation rules
Design Details
- →Data-driven claims: Claims generated automatically from enrollment data, eliminating manual entry errors
- →Rule validation: Every claim validated against eligibility and formatting rules before submission
- →Transparent feedback: Staff see exactly what rules apply and why a claim was approved or flagged
- →No manual submission: Claims submitted automatically with complete accuracy and tracking
Impact
80%
Faster patient enrollment
This improvement came from solving the right problems with better design.
Eliminated manual entry
Structured intake forms replace unguided data collection
Real-time eligibility
Staff get instant feedback instead of waiting for answers
Automated claims
Claims submitted automatically without email delays
Transparent workflow
Staff know exactly what step comes next and why
The speed improvement is a byproduct of better design, not the goal. The goal was making the process work.
What I Learned
User interviews beat assumptions every time. We thought the problem was complex. It was simple: nobody had built software for this before.
Process Insights
- →Iteration on real product beats pixel-perfect mockups
- →Design and code working together solves problems fast
- →User validation early and often prevents building the wrong thing
- →Small, shipped features beat big, unfinished ones
- →Talk to end users, not just stakeholders
Want to build something that actually works?
I focus on understanding real problems and shipping real solutions.
Let's talk