AI-powered mobile product designed to reduce doom-scrolling through reflective content and adaptive recommendation logic.
MindScrolling is a mobile-first philosophical discovery product built with Flutter, Fastify, and Supabase. Instead of optimizing for addictive engagement, it explores how recommendation systems can support intentional reflection.
Most feeds optimize for attention capture. MindScrolling explores the opposite: a feed designed to encourage reflection, curiosity, and healthier digital behavior.
- Reflective scrolling instead of addictive scrolling
- Behavioral recommendation instead of pure click optimization
- Mobile-first product experience
- Optional premium model without ad dependence
Flutter App
↓
Fastify Backend
↓
Recommendation Engine
↓
Behavioral Signals + pgvector Similarity
↓
Supabase / PostgreSQL
- Adaptive philosophical feed with 13,000+ quotes
- Swipe-based behavioral signals
- Philosophy map for content exploration
- Multilingual support (EN/ES)
- Premium packs / monetization layer
- Challenge engine for user engagement
To support a unified mobile product across Android and iOS with a single codebase.
To keep backend APIs lightweight, modular, and high-performance.
To personalize the feed using interaction patterns (swipes, likes, saves, reading time) rather than pure click-through rates.
| Layer | Technology |
|---|---|
| Mobile | Flutter / Dart |
| Backend | Node.js / Fastify |
| Database | Supabase / PostgreSQL |
| AI | pgvector + EMA hybrid recommendation |
| CI/CD | 14 pipelines |
| Distribution | Google Play Store |
backend/ # Fastify API
flutter_app/ # Flutter mobile app
docs/ # Architecture docs
data/ # Quote datasets
store-assets/ # Play Store materials
scripts/ # Build & deploy scripts
testers/ # Test utilities
| Metric | Value |
|---|---|
| Commits | 140+ |
| Platforms | Android / iOS |
| Public Releases | 3 |
| Quotes | 13,000+ |
| CI/CD Pipelines | 14 |
cd backend
npm install
npm run devcd flutter_app
flutter pub get
flutter run- Better recommendation explainability
- Offline caching improvements
- Feed analytics dashboard
- Reflection notifications
Built by Christian Hernandez · AI Engineer