OVERVIEW
Doctronic is an AI health platform with 25M+ consultations — built on a multi-agent LLM system trained exclusively on peer-reviewed medical data. I joined as Lead Designer to own the patient-facing product experience across AI chat, clinical workflows, and the company's most ambitious feature: an AI-powered prescription refill flow that would become the first of its kind authorized by a US state.
My ROLE
I drove design and UX including information architecture, interaction design, and visual design. I worked directly with the founding team, ML/AI engineers, and licensed physicians to define how LLM outputs should surface to patients at each step including:
The AI prescription refill flow (the primary conversion path)
The decision routing system (refill → AI consult → video doctor)
User journey optimization across the broader platform to improve engagement and CTR
UI refresh to align with trust, clarity, and clinical credibility signals
The Problem
Getting a prescription refilled is one of the most routine healthcare needs — and one of the most broken experiences. Patients face hold music, appointment backlogs, and unnecessary office visits just to renew a medication they've taken for years. Doctronic's technology could solve this, but the design challenge was harder than the engineering one:
How do you make an AI making a clinical decision feel safe, clear, and trustworthy to a patient who may be scared, confused, or already frustrated with the healthcare system?
Key Design Challenges
1. Designing trust for a high-stakes AI decision
When an AI is making a clinical recommendation — even a routine one — the interface has to earn trust at every step. I designed a transparent step-by-step intake flow (Find Prescription → AI Health Consult → Pharmacy Selection → Review & Payment) that mirrors how a physician thinks, not how a form works. Each step builds confidence rather than creating anxiety.
2. Routing users intelligently without friction
Not every user can get an instant AI refill — some need a video doctor, others need a full consult. The routing logic is complex on the backend, but the UX had to feel seamless. I designed the decision branch to feel like a natural conversation, not a hard rejection — users who couldn't get an instant refill were guided to the next best path without dead ends.
3. AI output legibility
The LLM surfaces clinical reasoning that informed doctors understand intuitively. For patients, that same information needs to feel reassuring, not clinical and cold. I developed interaction patterns for how AI-generated content, confidence states, and safety check confirmations should be displayed — building a visual language for AI transparency that the team could apply consistently.
Outcome
Shipped the first AI-authorized prescription refill system in the United States, in partnership with the Utah Department of Commerce
Featured in Washington Post, Politico, and Salt Lake Tribune at launch
UI and journey optimizations drove measurable improvements in engagement and CTR across the platform (add your specific % lifts here if you have them)
The refill flow now covers 190 medications and has processed refills for patients across all 50 states
What I Learned
Designing for AI isn't about making the AI visible — it's about making the human feel in control. Every interaction pattern I built for Doctronic was rooted in one question: does this make the user feel more confident, or less? That instinct now informs how I approach every AI product problem.