AI-Powered Dining Platform + Native Mobile App

Kitchen staff needed to publish daily menus quickly, but writing descriptions and finding images was tedious. Students wanted instant access to menus without hunting through PDFs or websites.

I built two systems that work together: a web platform that uses AI to generate menu content, and a native iOS app that puts menus directly on students' home screens and watches.

AI Content Generation (Web Platform)

The Problem: Staff type "Pan-Seared Salmon" but need marketing copy and an appetizing image.

The Solution: Integrated OpenAI's GPT-4 and DALL-E. Staff click one button and get polished description + hero image in under 30 seconds.

[Screenshot: Menu editor with AI generation button]

Impact: Eliminated manual copywriting and stock photo hunting. Menus published faster during busy periods.

Widget-First Mobile Experience

Most apps treat widgets as an afterthought. I designed the iOS app with widgets as the primary interface—students never need to open the app.

[Screenshot: Home screen widget showing today's menu]

Features:

  • Home screen widget with tap-to-toggle meal periods
  • Lock screen widgets for glanceable info
  • Apple Watch complications for wrist access
  • All interactive—no app opening required

Offline-First Architecture

Students check menus in underground lecture halls with zero connectivity. The app aggressively caches data using shared storage between the widget, Watch app, and main app.

Result: Menu loads in <50ms from cache, even with airplane mode enabled. Works 100% of the time, regardless of connectivity.

[Screenshot: Menu loading instantly offline]

Apple Watch Standalone App

Built a complete Watch experience optimized for the tiny screen and severe memory constraints (256MB). It fetches data independently of the iPhone and includes multiple complication styles for different Watch faces.

[Screenshot: Watch app UI + complications]

Digital Signage Integration

The same web platform drives 4K TV displays in dining halls. Built a clean iframe system that scales typography dynamically to any screen size and auto-refreshes every 15 minutes.

[Screenshot: Menu on dining hall TV]

The Complete Flow

1Kitchen Staff → AI generates content → Database → JSON API → iOS App → Widget/Watch

One system, multiple touch points, zero manual content creation.

Technical Approach

Backend: Laravel 12 with Livewire 4 for real-time collaborative editing. OpenAI API integration for GPT-4 and DALL-E. Dynamic viewport scaling for digital signage.

Frontend: Native Swift/SwiftUI leveraging WidgetKit, App Intents (iOS 17), and WatchOS. Shared data layer using App Groups for instant widget updates.

This demonstrates full-stack capability—from AI integration on the backend to platform-specific features on mobile.

← Back to Home