*Technical Architecture:* - SwiftUI frontend with offline-first design - Firebase for real-time leaderboards and user sync - Custom question difficulty algorithm based on historical performance - Edge computing for global question distribution at midnight local time
*Interesting Problems Solved:*
1. *Question Quality at Scale:* Built a content pipeline with fact-checking, difficulty scoring, and A/B testing. Questions that score <60% or >90% get recycled.
2. *Global Synchronisation:* Distributing identical questions worldwide at optimal local times while handling time zones and edge cases.
3. *Leaderboard Performance:* Real-time rankings for 30+ users updating simultaneously. Ended up with a hybrid approach using Firebase + local caching.
4. *Difficulty Balancing:*Custom ML model that predicts question difficulty based on category, length, answer type, and historical data.
*Unexpected Learnings:* - Content creation was 90% of the work, coding was 10% - Users care more about question quality than anything else - Daily engagement patterns are fascinating. Huge spikes at morning, lunch, and after 9pm - International users have very different knowledge gaps (geography questions vary wildly by region)
*Current Architecture Handles:* - 1k+ daily API calls - Real-time score updates for 30 concurrent users - 15,000+ questions with metadata and analytics - Sub-100ms global leaderboard updates
Happy to dive deeper into any of the technical aspects. The intersection of content, community, and real-time systems has been fascinating to build and learn from
Live on App Store: https://apps.apple.com/gb/app/qwiz-daily-quiz-trivia-game/id...