Users often receive health and assessment reports as dense PDFs that are hard to understand or act on. At the same time, AI tools either provide generic answers or lack a structured way to turn insights into personalized actions.
The challenge was to design an experience that:
Helps users understand complex health reports instantly
Allows users to ask questions naturally, without medical or technical jargon using our AI assistant
Transforms scattered information into clear, personalized protocols
Flows that work both for self-use and for creating protocols for others
I owned the design of the core AI experience, from understanding inputs to generating actionable outputs.
My work included:
Designing the AI chat assistant, allowing users to:
Ask general questions
Ask questions directly about uploaded PDF health reports
Designing the PDF upload and summarization experience, turning long reports into clear, readable insights
Creating the AI protocol generator, enabling users to generate protocols in three ways:
Using only a health report
Using only a text prompt
Using a combination of report + prompt
Defining clear user flows for creating protocols for yourself or for someone else
Structuring AI responses to feel trustworthy, actionable, and easy to follow
Establishing foundational usability principles for AI-driven interactions
Collaborating closely with founders, product, and engineering to align AI capabilities with real user needs
