Back

Product / UX Research · Internship · WellFit Pay · 2026

Designing Payment Arrangements patients can see.

WellFit is used in 1,000 dental offices, and this is the part that kicks in when a patient can’t afford their treatment. Only designer, 9 weeks, interviews to handoff.

Role
Product / UX Research Intern · sole designer
Industry
Health tech · Dental payments
Timeline
9-week internship · Summer 2026
Team
1 designer (me) · 2 product managers
Presented to
CEO, CFO and CPO
4features I designed end to end, solo, in 9 weeksHow it was measuredThe Propensity-to-Repay score, the patient dashboard (mobile + desktop), automated reminders and the office metrics dashboard.
78%average rating for the patient dashboard in pre-launch testingHow it was measuredAn average rating of 3.9 out of 5 from the pre-launch usability testing I ran, shown as a percentage. It’s a usability rating, not a business result.
76%average usefulness rating for the Propensity-to-Repay scoreHow it was measuredAn average usefulness rating of 3.8 out of 5 from the same testing, shown as a percentage.

Overview

About WellFit.

WellFit is a platform for dental offices that puts patient financing, dental plans and payments in one place, and it’s used in 1,000 dental offices.

I worked on WellFit Pay, the part that lets an office offer a payment arrangement when a patient can’t afford treatment and financing says no.

The problem

A payment arrangement was just a handshake.

WellFit’s goal is to never turn a patient away, but the arrangement they get comes with no contract. Here’s how a patient ends up in one:

The real problem

Digging in, it was never really about the handshake.

Problem 01Trust

The front desk had zero tools to decide who gets a plan, and nothing binding once they said yes.

Problem 02Miscommunication

Patients didn’t know what they owed, and offices tracked plans by hand (sometimes on sticky notes!).

Problem 03Late payments

Payments slipped, A/R days went up, and nobody reached out until collections.

Research

So I went and talked to everyone.

The business9interviews

Stakeholders across four teams

The workflow3offices

Dental front desks, online and in person

The market9platforms

None of them helped after a decline!

The market

The 9 platforms I looked at. Hover or tap one to flip it.

Financing at checkout 5

Practice payments & tools 4

After a decline? Nobody.WellFitPay✓ Built for right after a decline

What I found

Nobody could see the plan, not the office and not the patient.

Insight #1 · Trust60 moPlans were handshake deals, with no end date.They are doing handshake deals and they are not getting paid.CPOWhere it comes fromMy stakeholder interviews at WellFit: arrangements ran “60 months... with no cap in sight.”
Insight #2 · MiscommunicationBy handNo single place to see every plan.Staff were tracking agreements manually, sometimes on sticky notes.Dental officeWhere it comes fromMy visits to 3 dental offices, where front desk staff tracked agreements by hand, sometimes on sticky notes.
Insight #3 · Miscommunication40–50%of support tickets were basically “what do I owe?”Where it comes fromWellFit’s support data: 40–50% of support tickets were patients asking what they owed or when a payment was due.
Insight #4 · Late payments56%of practices saw A/R days go up. Nobody reached out until collections.Where it comes fromMy discovery research at WellFit: 56% of practices reported rising accounts receivable (A/R) days and manual follow-up.

Ideation

Want to see the first sketches? Yes, it was messy.

My manager JT and I drew out how every single feature would flow (detective evidence board vibes).

every way in (what if they fail P2R??)
the calculator (yes, my handwriting)
different room, same energy

Scoping

Then I narrowed 13 ideas down to 4.

13feature ideas from research
  • Propensity-to-Repay score
  • Patient dashboard
  • Automated reminders
  • Office metrics

The 4 that made it each go straight at one of those three problems: trust, miscommunication or late payments.

The features

4 features, one patient.

Propensity to Repay, the Patient Dashboard, the Office Dashboard Redesign and Automated Reminders. Follow one patient through all four.

Feature 1 · Front deskPropensity to Repay Feature 2 · PatientPatient Dashboard Feature 3 · OfficeOffice Dashboard Redesign Feature 4 · OfficeAutomated Reminders

Feature 1 · Front desk

Propensity to Repay

Should we even offer this patient a plan?

The problem. A repayment model existed, with no interface.

What I designed. After a decline, the system calculates a 1 to 10 score on its own, and it helps the office decide instead of deciding for them.

Fixes trust. The front desk isn’t guessing anymore, there’s a score behind the “should I?” call.

WellFit Pay prototype: the repayment score screen
  1. 1
    1 to 10, not a credit scoreAnyone gets it, whether you’re 5 or 80
  2. 2
    Three bands, three colorsRed is nah, yellow is maybe, green is let’s go
  3. 3
    The office still decidesThe score supports the call, it doesn’t make it
  4. 4
    One clear next stepOffer a plan, or end the session
Breaking down the score the front desk sees.
Design decisionThe patient never sees the score

Money talk at the front desk is uncomfortable, so the score stays staff-only.

What the front desk sees
3/10Low
What the patient sees
Remaining balance$1,584Next payment · Aug 15

No score, no credit number

Low score? A manager’s override code still lets the office offer a plan.

03Manager overrideStaffPlans beyond standard approval limits need a manager’s code.
Interactive demo

A patient’s financing just got declined. Watch the score come back.

Staff-only view · sample patients

Anna A.Financing declined
Calculating repayment score…
3/10Low
1–4 Low5–7 Medium8–10 High

Try itHit “Next patient” to run another score, then offer a plan.

Feature 1 · Once they say yes

Then the office sets up a plan that actually ends.

The problem. Plans never ended, and payments ignored payday.

Fixes trust and late payments. A plan with a real end date is way easier to trust than a handshake. And putting due dates on payday makes paying on time a lot easier.

Open-ended, up to 60 months4 months to 3 years
One due date for everyoneDue on the patient’s payday
SSN / ITIN at every stepEntered once
04Payments on paydayStaffFrequency and day of week match when the patient gets paid.
05A dated scheduleStaffEvery payment gets a date, then the patient gets their dashboard.
Interactive demo

Now build Anna’s plan with the new rules

Illustrative · $1,584 balance from the patient dashboard

Term length (set options)

Payment frequency (matches pay schedule)

$0

Before: arrangements stretched to 60 months with no ceiling.

Try itPick a term length and a pay frequency.

Feature 2 · Patient

Patient Dashboard (mobile and desktop)

What do I owe, and when?

What I designed. A dashboard patients open from a text, on their phone or their computer: balance, next payment, and pay right there.

Fixes miscommunication. No more calling the office to ask “what do I owe?” since it’s right there (and they can pay on time from the same screen).

Testing. My strongest score, 78% (“very easy and user friendly to navigate”).

Low-fi recreation. Pay on the computer, and the phone shows it too.
Interactive demo

Be Anna for a minute: check the balance, then pay it

Rebuilt from the tested prototype

Wellfit Pay
ABC Dental Group, ...ABC Anaheim DentalGU
APW DentistryDashboardPaymentsHistoryContractReturn to Menu
2
Anna AllowedAPW Dentistry

Remaining balance

$1,584

34%

$816 paid$2,400 total

Remaining balance

$1,584

of $2,400 total

Powered by wellfit

$816 paid$2,400 total

Overdue

$105

1 payment past due
Return to Menu

Try itOpen the Remaining tile, then make a payment or clear the $105 overdue. It’s pretend money, promise.

Feature 3 · Office

Office Dashboard Redesign

Who’s behind, and by how much?

What I designed. A new office dashboard, and the big add was four metric cards up top. Tap one and it drops down to show exactly who’s behind the number.

Fixes late payments. The office sees who’s overdue the moment it happens, instead of finding out at collections.

Interactive demo

Tap a card to see who’s behind the number

Rebuilt from the tested prototype

Overdue accountsTap a card to see details

AccountAmountDays overdue
Smith, John$42545 days
Johnson, Maria$38032 days
Williams, Robert$51028 days
Davis, Sarah$43021 days
Brown, Michael$40015 days

14% of total receivables are past due

Past due (14%)Current (86%)

43 accounts past due out of 307 total accounts

Payment arrangements by status

  • Active28
  • Pending8
  • Completed5
AccountAmountDays
Miller, James$21060 days
Wilson, Patricia$34045 days
Taylor, Robert$18038 days
Thomas, Linda$29018 days
White, Elizabeth$41518 days
Martin, William$31012 days
Try itTap any of the four cards to drop it down (and again to close it).

Feature 4 · Office

Automated Reminders

How do we stop payments from going quiet?

What I designed. Reminder rules the office sets once: a heads-up before a payment is due, and a nudge after one is missed.

Fixes late payments and miscommunication. Patients hear about a payment before it’s late, so nothing goes quiet until collections.

Interactive demo

Set Anna’s reminders, then save them

Rebuilt from the tested prototype

Payment reminders

Days before payment to send reminder

Reminder message preview

“Hi Anna, your payment of $85.00 is scheduled for Aug 15.”

Past due reminders

Days after a missed payment to send the first reminder

Past due message preview

“Anna, we noticed your payment didn’t go through. Tap here to fix it.”

Try itChange the days, then hit save.
What testing saidStaff wanted proof it saved

Reminders scored lowest (72%): staff weren’t sure their settings saved. Lesson: design the proof, not just the setting.

Testing

Then I put it in front of 75+ people.

Before launch2rounds

Planned and run by me

Unmoderated75people

On UserTesting.com

Live usability testingStaff+ an office

WellFit’s staff and a partner dental office, online and in person

Scores are average ratings as a % (3.9 out of 5 = 78%). No critical accessibility blockers.

78%

Patient self-serve dashboard

“Very easy and user friendly to navigate.”

76%

Propensity-to-Repay usefulness

“Would save time by reducing individual calls.”

72%

Readiness to switch on reminders

“Staff wanted proof that settings actually saved.”

Outcome

And the CEO approved the designs and wants to push this out.

Pitched in person to the CEO, CFO and CPO, then handed off all 4 features plus the roadmap.

  1. 19 interviewsStakeholders across four teams
  2. 23 officesFront desks, online and in person
  3. 313 → 4Ideas scoped down to 4 features
  4. 475+ peopleTested before launch and live
  5. 5PitchedIn person, to the CEO, CFO and CPO
  6. ✓ApprovedThe CEO approved the designs and wants to push it out
The WellFit team

Huge thanks to Karey Crump, Johnathan Tebeau and Paul Little for trusting me with real ideas from day one.

“He took a messy problem and had a working prototype in 9 weeks. His execution was phenomenal. He had incredible drive and ambition in creating this product. Would highly recommend Sahil for any role.”
Johnathan TebeauProduct Manager, Wellfit Technologies
“I gave him one of our underutilized payment products and asked him to figure out why it wasn't landing. What he brought back was a rethink of the whole strategy, built around what would actually get healthcare offices to adopt it, along with a ranked list of enhancements by impact and effort. Any team would be lucky to have him.”
Paul LittleProduct Manager, Wellfit Technologies

Key learnings

What I’m taking with me.

This was my first time with a huge responsibility: the solo designer, presenting features to internal stakeholders.

It was tough at first.Honestly, I didn’t know where to start. Now I’m way more confident presenting to a room.
I had tremendous support.JT, Paul and Karey kept me motivated the whole way.
Know who you’re talking to.Users cared about their tasks, PMs cared about priority (13 ideas down to 4!), and the C-suite just wanted to know people would actually use it.
Figure out what success looks like first.Making pretty UIs is honestly the easy part. Next time, I’m locking in the metric on day one.

This is just a snapshot! Want the long version? Grab 30 minutes with me.