Challenges & Solutions

• Building an intuitive 3D body model for pain point placement on mobile╰ SceneKit with quaternion-based trackball rotation, axis-locking, and haptic feedback
• Keeping pain data in sync across iOS and web with offline support╰ Offline sync queue with automatic retry using Network framework, Supabase for real-time sync
• Migrating the web app from client-side to server-side rendering╰ Hybrid RSC + client component architecture, modularized data layer for tree-shaking
• Designing a doctor-patient workflow with role-based access╰ Supabase Auth with invite flow, RLS policies, and server-side permission checks

Technical Details

iOS app uses SceneKit for 3D pain point placement on a human body model. Web app built with Next.js using server-side rendering, Three.js for 3D visualization, and Supabase for auth and data. Express API serves as middleware between iOS and Supabase. Hosted on Cloudflare Tunnel.

Contributions

iOS

• 3D body model with SceneKit — trackball rotation, axis-locking, sphere placement/editing
• Full UI redesign following Apple HIG (auth, tabs, pain log, history, settings)
• Offline sync queue with automatic retry
• Architecture refactor to proper SwiftUI patterns

Web

• Next.js app with SSR migration from client-side rendering
• Three.js 3D pain visualization
• Sidebar navigation, theming, doctor invite flow
• Modularized data layer and API cleanup
• Database schema design for pain sessions

API

• Express server with session/user routes and auth middleware

Infrastructure

• Cloudflare Tunnel hosting
• Supabase configuration and RLS policies

Learnings

• SceneKit 3D interaction — quaternion rotation, gesture handling, sphere placement
• SwiftUI architecture patterns — @EnvironmentObject, dependency injection
• Next.js SSR migration — hybrid RSC + client components
• Three.js 3D visualization in React
• Supabase Auth with role-based access and RLS policies
• Offline-first patterns with sync queues

Future Plans

• Revive as a simplified logging-only app