I'm an iOS developer based in Brisbane, Australia, building Swift & SwiftUI apps across health, fitness, and productivity. I have 5+ years of experience shipping apps to the App Store and a strong interest in on-device AI and ML.
- 🏆 2nd Place – Google MedGemma Impact Challenge (March 2026) among 850+ global teams
- 📱 Multiple live apps on the App Store
- 🤖 On-device ML with Apple MLX, CoreML, and Gemini AI
- 📺 iOS development content on YouTube @developerjosh
- 🌐 joshbourke.com
Social fitness app with real-time step leaderboards powered by HealthKit and Supabase · steppy.fit
- Friends compete on daily step leaderboards with live HealthKit sync
- Custom Supabase backend with RLS policies, PostgreSQL RPCs, and Edge Functions
- Group management, username validation, reporting system, and App Store-compliant account deletion
- Profile customisation with SF Symbol icons and a SwiftUI parallax hero view
Stack:
SwiftUIHealthKitSupabase@Observable
AI-powered whole food nutrition tracker · nutrify.app
Kaggle competition-winning iOS app for AI-powered home inventory · keeptrack.app
- 🥇 Kaggle competition winner
- Gemini AI analyses photos of household items to autofill name, brand, description, location, and estimated value
- Firebase backend for cloud storage and sync; supports receipt/warranty attachments and PDF export
Stack:
SwiftUIGemini AIFirebase
2nd Place – Google MedGemma Impact Challenge · github.com/mrdbourke/sunny
🏆 Recognised as 2nd place in Google's MedGemma Impact Challenge (March 2026) among 850+ global teams — featured on the Google Keyword blog.
- Built the iOS app for an open-source on-device skin health tracker
- Fine-tuned MedGemma-1.5 on 1,000+ images; deployed on-device via MLX 4-bit quantisation
- Generates structured skin reports from photos while keeping all data completely private on-device
Stack:
SwiftUIApple MLXMedGemmaCoreML
