miterun
Real-time occupancy counter for cafes, barbershops & small stores π’π‘π΄
About miterun
miterun (γγγ«γ³) is a lightweight PWA that lets small shops show their real-time occupancy to customers. Point a spare smartphone at the room and it counts people on-device with TensorFlow.js (COCO-SSD MobileNet v2) β only the head count is sent, never video β while staff can override the status with one tap. Customers check a public page before they head over, so cafes, barbershops, clinics and small stores stop losing walk-ins to "is it crowded right now?" uncertainty. It also handles reservations with SMS and email reminders, push and LINE alerts when a store fills up, weekly congestion analytics with AI wait-time prediction, and Square/Stripe POS correlation. Built on Cloudflare Workers, multi-tenant, installable to the home screen, works offline, and takes about five minutes to set up with no credit card.
Key features
- On-device AI people counting with TensorFlow.js
- One-tap manual status override for staff
- Public occupancy page customers can check before visiting
- Reservations with SMS and email reminders
- Push and LINE alerts when a store fills up