Back to home
How can we help unpaid family caregivers spend less time as a caregiver, and more time as a daughter, son, or spouse?
Role
Senior Product Designer
Contribution:
Visual design, UX, content and product strategy
Team:
Health Futures, Microsoft Research
Duration
1.5 years
Platform:
web app
Tools:
1 in 5 Americans are caregivers for aging parents or spouses, often relying on general productivity tools like Excel, Word, and Teams to manage medications, health information, and appointments. These tools, while functional, aren't tailored to the unique challenges caregivers face, leading to inefficiencies and added stress. The objective of the Health Futures team was to develop a dedicated app that streamlines caregiving tasks, allowing caregivers to spend more quality time with their loved ones and less time on administrative duties.
When I joined the team, it felt like a mini startup within Microsoft Research. We were truly 0-1, starting from scratch. With newly approved budget and headcount, we had the opportunity to move from early exploration into real product development.
Our task was to take the white paper and early technical prototype, collaborate on user research with our PM, and rigorously scope it down into a focused MVP that balanced technical feasibility with what caregivers would truly value.
Our team
I was the Senior Product Designer on a scrappy, startup-like cross-functional team that included PMs, researchers, and engineers
Role & Goal
The business goal was to explore new opportunities in health and caregiving under Microsoft Research and identify AI-driven consumer applications.
My role spanned
Early research synthesis
UX Design
Prototyping
MVP development
Leading and hosting hackathons
Key Results
Solved a complex medication flow
Simplified how users track and manage medications by redesigning the calendar experience—based on deep empathy interviews and behavioral mapping.
Reduced med errors with LLMS
Prototyped an AI-driven interpretation system using Universal Medication Standards—built during a hack week to clarify confusing dosage instructions.
Led clinical design (as the most junior designer)
Promoted to own all clinical UX—collaborated with Teladoc & Epic after standout work on the medication feature.
Led the Caregivers Council
Coordinated workshop sessions to review prototypes, gather feedback, prod into needs and insights
Drove AI strategy post-GPT-2
Ran a company-wide hackathon to explore AI/LLM integration—sparked a new product direction by connecting tech potential to user journeys.
Transformed cold UI into a lovable product
Partnered with visual design to evolve a sterile Microsoft Fluent base into a warm, accessible health tool that exceeded a11y standards.
"Sally played a pivotal role in leading the concept-to-creation phase of several critical features of our caregiver platform. Her ability to collaborate effectively across diverse teams, including design, engineering, product management, and marketing, is truly commendable. Sally's leadership in hosting workshops with users and stakeholders to generate innovative concepts and envision core product features was instrumental in our platform's success."
Suff Syed, Head of Product & Design
Microsoft
Find the right problem
Caregivers juggle countless responsibilities…Which are the most urgent to solve for in our MVP?
Become an expert, fast
I became an expert in family caregivers as soon as possible by synthesizing every interview conducted by our PM, familiarizing myself with our 42 page white paper, and reading through posts on facebook and reddit to grasp perspective
Insight: Unpaid family caregivers spend approx 24 hours/week on care, face twice the depression rate of the general population, and nearly 1 in 4 report reduced work productivity—costing employers $5,600 per caregiver annually.
Translate & understand user needs
We developed a comprehensive set of "user can" statements, and then validated and prioritized them through surveys, card sorting, usertesting.com and must critically, through our internal Caregivers Council.
I ran ongoing design sessions asking questions like:
What could a memory feature look like?
How do you organize medication lists?
Who helps you make medical decisions, and how?
Our 3 feature sets aka our "Epics." Not pictured: Calendar, home, and settings.
Win: We launched the Caregivers Council—a panel of Microsoft employee caregivers who I hosted bi-weekly and engaged in a teams channel to aggregate feedback, run polls, and prod into pain points. It hugely accelerated our velocity in both research and design.
Ruthlessly prioritize
We relied on the Kano model to identify "must-haves" vs. "delighters."The results were telling: features related to medication management (Medication List, Refill List, and Medication Schedule) ranked high in both satisfaction and dissatisfaction. This positioned them as clear performance drivers.
Kano model framework
Medication list ranking #1 in Importance to our users
Our prioritization
Insight: We interpreted this as a signal: caregivers would only trust and use the app if it solved medication stress first. This led us to assume that medication tracking should be our core MVP focus.
Build the right thing
How can we ensure Medication manager delivers real value across use cases by balancing what caregivers need, what's technically possible, and what's easy to use?
Artifacts of the new Add Medication flow I designed
Just enough research
To understand how to design a medication schedule I intervied 15 family caregivers caregviers, analyzed the top conditions they are caring for (Hypertension, High Cholesteorol, Arthritis, Alzheimer’s, Parkinson’s, Diabetes, etc), and then reviewed sample med lists for each of these conditions with them. I learned how complex med schedules are, which helped me design a flow that reflects real med types and their timings.
Insight: 50% of caregivers manage 5 or more medications per day for their loved ones. 20% manage 10+ meds.
I used this file to QA my designs and ensure it fit these common but complex medication schedules
Audit the v1
I reviewed earlier mockups from a contracted designer. With our validated feature set, it became clear the design didn’t scale for real-world needs. Key features like flexible dosing, medication types, and smart scheduling were missing
Collect pain points, opportunities, insights
Medication schedules
Caregivers often struggle with tapering instructions, “as needed” meds, and dosage changes that vary daily
We can autogenerate common medication schedules. For example, if a user added a common antibiotic "z-pack" we can auto suggest the default schedule for the user to review.
Medication schedules are super naunced, will be a challenge to simply in the UX, but this data is important, especially if we want to integrate with EHR in the future |
Adherence vs Reminders
Caregivers were less interested in tracking adherence manually (whether meds were taken) and more on sending reminders to their loved ones. |
Prioritize reminder systems over adherence checklists |
|
We experimented with having users reply Y or N to mark taken, but ultimately nixed the complexity for now so users did not need to mark each med taken. We were also exploring Alexa & Smart TV notifications.
Symptom tracking
Caregivers aren’t just logging meds—they’re trying to track what's working, what’s not, and if symptoms are changing. |
Combine medication tracking with symptom and side effect monitoring to surface patterns over time. |
Tracking meds is less about checklists, more about ensuring safety and speaking confidently to providers atappointments. This is an important connection point we can tap into.
Innovate the product
During a hack week, I partnered with our data scientist and engineer, Katie, to build an GPT2-powered system that translated confusing medication instructions into clear, user-friendly schedules. We based it on a system called the Universal Medication Schedule (UMS) that helps standardize dosing times and reduce errors—making it easier for caregivers to understand and follow complex regimens.
Dosing Instructions are confusing:
Why this is important
In one study of older adults, patients were given 7 medications and asked to create schedules.
Only 14.9% took medications at the proper dosing time (which was 4 times per day). Instead, almost a third of users scheduled dosing times greater than 7 times per day. When medicines had variable expressions of the same dose frequency (“every 12 hours” vs. “twice daily”), 79.0% did not consolidate the medicines. Additionally, nearly half of patients misinterpreted common instructions (like take every 4-6 hours or twice daily) when attempting to dose out a single prescription medication.
Win: I utilized the design language of the UMS symbols and colors in our design.
Utilize AI
I lead a hackathon with our entire team to understand how we could utilize Open AI's technology - which we were granted early access to as Microsoft AI employees. This ultimately led to several new feature innovations - the largest being carebot (a chat based interface where caregivers can ask questions about medical information in the app, and be served helpful information when they need it - ie, preparing for a doctor's appointment, or what to do after a fall.)
Build trust with engineers
I worked side-by-side with engineering and our data scientist while building the product, and also during our frequent "Hack weeks."
"Sally is a fantastic UX designer and user researcher. We worked together on a healthcare app that required complex data input to be easy and fast for the end users as well as ML integration to be transparent. Sally tackled these complex UX challenges with ease and baked user research into every step of her process. She thinks product wide and collaborates across roles to ensure high quality designs."
Katie Claveau
Senior Data Scientist
"Sally was the beating heart of our team, and her contributions to our organization were invaluable. Her talent and work ethic are unparalleled, and her passion for design is evident in all her work. I strongly endorse Sally and encourage any organization looking for a top-notch designer to consider her for their team. She will undoubtedly bring creativity, value, and a sense of purpose to any project she works on."
Alex Garcia
Senior Software Engineer
Finding the right problem
