Crave
AI-driven restaurant management system: a customer ordering app, a website, and a desktop POS and back office.

Discover: swipe deck of dishes, reordered by the taste profile1 / 13
The 24-second film
The product
Crave is a complete restaurant product built as a SaaS: a white-label customer app, a marketing website, and a point-of-sale and back office that installs like a normal desktop app. The customer app treats deciding what to eat as the product, with a taste profile that learns from every swipe and re-ranks what each guest sees. The back office runs orders, tables, kitchen tickets, menu, stock, staff and reports on Supabase, with AI-driven sales and stock insights.
Highlights
- Taste-profile engine: every swipe updates top vibes, spice tolerance and average spend, then reorders the Discover deck and biases recommendations, all on-device.
- Budget Builder searches every main, side and drink combination and scores them against the budget and the guest's taste profile.
- Group ordering with per-guest tagging and bill split, dine-in send-to-kitchen, loyalty points and a live kitchen feed.
- White-label by design: brand, palette, menu, stores, promos and loyalty rules live in one config file, so the app re-skins per restaurant.
- Tauri 2 desktop POS and back office with native installers for Windows, macOS and Linux, in-app auto-update, PDF bills, Excel exports, and English and Urdu UI.
Status and next steps
In testing. Feature-complete, being hardened before launch. Connect the customer app to the live back office (it runs on sample data today), then code-signing and launch to the first restaurants.
My role
Platforms
- Android
- iOS
- Web
- Desktop
Stack
- Flutter
- Tauri 2
- React 19
- Supabase
- TanStack Query
Status
In market
