Patient Health App

Clarity is an AI-powered health companion I built to help users navigate their health journey end-to-end. Starting with a conversational chatbot, users chat with a health companion to ultimately arrive at a ranked differential diagnoses, then get guided through understanding their insurance, finding and managing appointments, and preparing for their doctor's visit.

Last summer, I discovered a tumor on my foot and spent hours chatting with ChatGPT to figure out what it was and whether I should be worried. Furthermore, having never navigated the US health system before, I struggled to understand my insurance, schedule the right appointments, understand medical notes and reports, and even know what questions to ask my doctor.

I used various tools to get through it such as ChatGPT, ZocDoc, Insurance Portals, Excel, and my iPhone Notes app. I built Clarity as the single tool I wish I'd had, and think others would find it just as useful.

Clarity is a React + Flask web app backed by Supabase and deployed on Vercel.

• AI Diagnostic Chat: The Claude API conducts a Socratic patient interview, gathering symptom details through focused follow-up questions before surfacing differential diagnoses ranked by confidence, with clear guidance to seek professional evaluation.
• Health Journey Manager: A persistent session tied to a health concern, moving through intake, diagnosis exploration, appointment booking, and follow-up, with state persisted across sessions.
• Insurance & Appointment Management: Users store their insurance details and log doctor visits directly in the app, keeping everything relevant to a health episode in one place.
• Tech Stack: React 19, React Router 7, Flask, Anthropic Claude API, Supabase (PostgreSQL + Auth), Heroku, Vercel.

Clarity provides the clarity that patients navigating the US health system need in a calm, guided process. It handles all of the hard work while keeping the user organized and guiding them through every step of getting care.