Loop Health — AI Assistant

Loop’s general-query AI assistant helps members get fast answers without leaving the app — while keeping trust, clarity, and safe escalation to human care.

The problem

Members had quick health questions but no lightweight way to get reliable answers inside Loop. Support and doctor flows were too heavy for “what does this mean?” moments.

Approach

I designed the assistant around progressive depth — short answers first, optional detail, and clear paths to book a doctor or talk to support when the query needs a human.

Conversation states — loading, partial results, cited answers, and error recovery each have distinct treatment so members always know what the system is doing.

SwiftUI prototyping — built interactive prototypes on device to test timing, keyboard behavior, and handoff to care flows with engineering early.

Tools

SwiftUI — native prototyping and production patterns
Figma — flow mapping and conversation UI specs