This initiative marked SmithRx's first venture into AI, designed to establish a foundational framework and strategy for intelligent design that enhances business efficiency.
As a pilot initiative, we developed an MVP to streamline Member Support Staff (MSS) workflows. After releasing the pilot to a small group of MSS agents for testing and feedback, I joined the project. I brought a usability perspective to guide our experience strategy and execution, focusing on user research and synthesis, product usability, and identifying new use cases.
Company
SmithRx
Year
2025
Role
Lead Product Designer
Skills
AI, user research, team alignment, prototyping, high-fidelity mockups
After piloting the MVP to internal users, our research uncovered 3 key learnings: 1. Information overwhelm 2. Content (dis)organization 3. System confusion
Although the content showed significant potential value, users felt overwhelmed by its poor visual hierarchy—creating a "wall of text" effect. The difficult-to-read formatting made it challenging to quickly identify key insights. Users also lost trust in the system and were reluctant to adopt it due to confusion about its capabilities and data refresh rates.
Based on our research findings, we developed a solution framework aligned with our AI design strategy: 1. Information overwhelm → Visual clarity 2. Content disorganization → Structured outputs 3. System confusion → System transparency
To scale the MVP, we expanded our user base by including the SmithRx Clinical team's Prior Authorization approval workflow.
Through discovery workshops, we aligned our current and potential product capabilities with opportunities to improve workflow efficiency, particularly in information retrieval and synthesis.
We continuously refine our strategy to deliver high-quality content while maintaining cost efficiency.
For example, we are exploring breaking down large chunks of information into smaller components to optimize model size while maintaining output accuracy and completeness.
Key project considerations
1. Long-term integration While this project currently exists as a standalone application separate from SmithRx's internal platform, our long-term vision is to fully integrate AI capabilities into our existing ecosystem. This integration presents complex challenges in both backend configuration and user interactions—challenges we are actively exploring as we work toward our North Star. 2. Design library Given the above, the pilot application was and continues to be built using a third-party design library, introducing a boundary to our sandbox of short-term UI iterations.
Credits
Next Project
