My Role
UX Researcher
Timeline
November 2023 - March 2024
Team
Maryna K., Lead Software Developer
Hyeonyoung P., Front-End Developer
Role of interpreters
〜25% of appointments require an interpreter
Since some physicians speak only English while patients speak Ukrainian or Russian, interpreters play a critical role in enabling communication and ensuring quality care.
problem
The availability submission process was inefficient and prone to errors
The form required interpreters to manually click on each time slot individually and double-check dates with an external calendar.
The process of adding availability
research
To better understand the problem, I surveyed 7 interpreters
The goal was to learn about their experience with the current flow and how they usually submit their availability.
Given that all team members are volunteers, a survey format allowed them to share feedback on their own time.
Key pain points reported
1
The day-of-week labels are absent, and the time slots are not visible when scrolling down, making the process more difficult.
2
Repetitive clicking to select/deselect individual slots.
3
Different processes for adding and removing availability.
4
Option to remove submitted slots only one by one.
In addition to identifying pain points, the survey also revealed valuable context that helped me shape the direction:
1
All interpreters reported using a laptop as their primary device to submit availability.
2
6/7 interpreters typically submit their availability in bulk for 2–3 weeks at a time and then update as needed.
These findings allowed me to prioritize a layout optimized for larger screens and showed me the need for a low-effort way to set recurring availability and make quick adjustments.
ideation
My goal was to minimize effort and cognitive load while ensuring data accuracy
I started the design process with a sketch to explore how a calendar-based approach could replace the repetitive, slot-by-slot form. I used it to present the concept to the team, align on the overall direction, and gather feedback early.
Inspired by scheduling tools like Calendly, the sketch introduced key elements such as:
1
A monthly calendar view to provide visual context and help interpreters align dates with days of the week.
2
A “Make me unavailable” checkbox to give interpreters a fast way to block off time.
3
A time range selector to reduce repeated clicks for common availability windows.
4
A dropdown to apply availability across recurring days, like all Tuesdays or weekdays.
usability testing & iteration
I translated the idea into low-fidelity wireframes and tested the prototype with 3 participants
Each interpreter was asked to walk through the process of setting their availability while thinking out loud, allowing me to observe their experiences.
The “Apply to” menu turned out to be a significant source of confusion.
The label was vague, and the interface didn’t clearly indicate which days were affected after making a selection, leading to confusion.
result
By redesigning the flow, we could decrease the number of steps by up to 76%*
With a faster and more intuitive process, interpreters can submit their availability with less effort and fewer errors, leading to more consistent and accurate scheduling. This minimizes scheduling gaps and improves the organization’s capacity to provide care for patients.
*Test scenario: submit a 3-hour availability for the next 8 Tuesdays:
Old flow's total steps: (3 steps × 8 Tuesdays) + 1 save = 25 steps
New flow's total steps: 5 steps + 1 save = 6 steps, which is 19 steps fewer
The final prototype
How might we prevent interpreters from accidentally removing booked time slots?
Interpreters receive an email notification every time a slot is booked, and it is also visible in their upcoming appointments dashboard. But what if they accidentally remove time slots from the " My availability" page that were already assigned to appointments?
To address this, I designed a system that prevents interpreters from making themselves unavailable on days with booked slots and requires them to contact a team to help with canceling or finding an alternative interpreter. This reduces the risk of last-minute cancellations and helps ensure a smoother experience for everyone - interpreters, patients, and providers.
reflection
One of the biggest lessons came from an unmoderated usability test that didn’t go as planned
I used Maze to gather early feedback, expecting quick insights, but the limited interactivity of the prototype caused confusion.
The misclick rate was 83%, and the average time spent was just 37 seconds. Participants assumed they were doing something wrong when elements didn’t respond on click.
“Somehow, I couldn't make myself unavailable… I did something wrong :(”
This experience reminded me about the importance of setting clear expectations in testing and ensuring prototypes behave as users expect.