Technology is improving, but infant mortality is rising - Leveraging AI for maternal healthcare.
About this project
Despite advances in technology, infant mortality rates in the U.S. have dramatically increased from 17.4 deaths per 100,000 live births in 2018 to 32.9 in 2021. In 2022, South and Midwest regions recorded the highest infant deaths in the US. There is a clear need for increased access to healthcare for pregnant mothers in underserved communities. I worked with University of Michigan's school of nursing to create an AI-powered platform to allow nurses to provide 1:1 coaching to mothers and caregivers without access in rural Michigan.
Role
Product design lead
Timeline
Jan 2025 - May 2025
August 2025 - ongoing
Team
4 engineers
2 data scientist
1 faculty mentor
Skills
Product design
UX Research
Stakeholder communications
Our initial user and secondary research confirmed 2 main barriers to effective maternal healthcare support.
The Stigma Barrier (Mother)
Fear of Judgment: Mothers avoid seeking crucial help due to fear of stigma, causing them to isolate and struggle with fragmented information.
The System Barrier (Nurses)
Workload Crisis: Nurses struggle with managing workload, limiting time and capacity for personalized coaching.
The goal
How might we reduce the clients’ stigma barrier and increase nurses’ coaching capacity to improve maternal education and reduce mortality risks?
The solution for mothers
An online maternal health platform with 1:1 coaching, personalized care plans and verified educational materials.
The solution for nurses
AI integration in nurses' workflow to reduce adminstative task and increase working capacity.
Our process
Our project was far from straightforward.
In this project, I had to design alongside the development rather than handing off a polished high-fidelity prototype. We relied on cycles of design, testing, feedback, and redevelopment, to balance technical feasibility with user needs and stakeholder expectations
The complete information architecture
An information architecture designed to cater to the unique needs of three distinct user groups
In designing this architecture, I needed to ensure that each user group’s needs were met. Clients shold be directed to educational content, appointment management, and personalized care plans, whereas nurses can access client data, surveys, and session notes, along with tools for creating and updating care plans; administrators can access and both nurse and client databases to efficiently assign nurse-client pairings.
Laying out the user journey to identify our opportnities
Pivoting from a high-risk feature to a unique value proposition
The initial plan was to use our AI engine, UM Maizey, for geospatial analysis to find local resources BUT from testing with prompt engineering, we found that Maizey can't reliably work with numbers and is not good with addresses. Based on the user journey map, I identified the necessary pivot to using the LLM for text-heavy care plan.
Major design decision #1
Design details
Incorporating AI into nurses workflow responsibly
I referred to IBM’s AI principles as guidelines to develop the care plan creation workflow. The design’s goal to help reduce nurse workload but still making sure the information provided is accurate and personalized.
Design for Mental Models
I used the care plan structure familiar to nurses, and inserted the resources finder in the implementation steps where nurses most likely to need it.
Provide Rationales for Outputs
The generated care plan links its content with sources such as meeting transcripts, notes, or survey results. User can hover over to see the quote or click on the annotation for the full document.
Design for Co-Creation and Variability
The generated texts are insert into text input boxes that users can click on and make direct edits that auto saves. Users also have the option to clear and regenerate text if needed.
Design for Appropriate Trust & Reliance
Notice of generated care plans are included at the top and bottom of the form.
A scheduling flow with structure and autonomy
Nurses can input their availability for nurse coaching using a grid schedule, and clients can later see this availability as timeslots for which they can request appointments.
Balancing different stakeholders scheduling needs
Nurses have different preferences when scheduling with clients
The solution for everyone
Nurses can set availability and recurring appoinments based on agreement with their clients
Nurses can go into editing mode to select and deselect the time slots they have available for nurse coaching.
Appointments can be set as recurring, and are displayed as time cards on the calendar.
Clients can see a list of upcoming appoinments, and reschedule based on nurses' provided availability for health coaching.
Clients' appointment list is displayed in a row for increased scanability.
Clients go through a 2-step cancel request process to avoid unnecessary cancellation.
Major design decision #3: Prioritizing the platform’s values
Focusing on the user needs and platform value proposition
The initial version of the client dashboard covered these features BUT considering the web-app constraints - users will only check in from time-to-time, I chose to priotize peronalized care-plan on the main dashbroad.
Impact & Contribution
I successfully led the project from a problematic scope to a strategically valuable, user-centered product, meeting the Expected Final Deliverable tier of success.
Design Value Realized
Successfully advocated for an expanded scope beyond the initial 5 screens, creating complete, functional workflows for 3 user groups and validating the role of UX as product strategy.
Strategic Clarity Achieved
Defined the final, feasible product scope by de-risking the project from a technically flawed AI strategy, leading to full stakeholder alignment.
User Confidence (Qualitative)
Early feedback indicated a high level of comfort, ease of use, and intention to participate in a nurse coaching program and receive personalized coaching. More research should be done about how to effectively inform the public about the program.
Nurse Efficiency (Projected)
Projected to yield a significant reduction in nurse administrative time and directly increasing capacity for high-quality, personalized patient care.
What I learnt from the project
I learned how crucial it is to start with well-defined stakeholder interviews and a strong UX focus from the very beginning. Having usability testing set up early also made it easier to test continuously and refine ideas instead of waiting until the end.
The biggest challenge was communicating the value of UX to a highly technical team. My proposed research and design timeline was met with pushback for potentially delaying development, but I learned how to advocate for testing as you go and show that good UX actually accelerates progress in the long run.