██████ ██████ ██████ ██ ███████ ██████ ████████
██ ██ ██ ██ ██ ██ ██ ██ ██ ██ ██
██████ ██████ ██ ██ ██ █████ ██ ██
██ ██ ██ ██ ██ ██ ██ ██ ██ ██ ██
██ ██ ██ ██████ █████ ███████ ██████ ██
██████ █████ ██████ ███████ ███████ ██ ██ ███ ██ ██████
██ ██ ██ ██ ██ ██ ██ ██ ██ ████ ██ ██
██ ███████ ██████ █████ ███████ ████ ██ ██ ██ ██
██ ██ ██ ██ ██ ██ ██ ██ ██ ██ ██ ██
██████ ██ ██ ██ ██ ███████ ███████ ██ ██ ████ ██████
We built CareSync during the Shipfast hackathon organized at Ecole 42. We wanted to make healthcare emergency department workflows smarter, faster, and more responsive using AI. We thus created an app that helped doctors fasten the patient priority sorting process. This sorting is actually made with the help of a long ten page paper grid (at least in France), and there are special nurses that are formed to set patient priority (they’re called the ‘triage nurses’). We thought this process could definitely be enhanced by a bit of computing and AI.
The application hinges on a React frontend styled with Tailwind CSS and Shadcn/UI, backed by Supabase services (PostgreSQL).
On the AI side, we integrated Anthropic’s Claude API for natural-language-based patient check-in and triage support, used Web Speech API for voice interactions, and added image analysis capabilities for medical photos.
→ Check CareSync here.
→ Check the repo here.
→ Check the creative side of the project here.