About the Project: tuSpeech

Inspiration

The idea for tuSpeech came from a simple need: I spend a lot of time reading long documents, articles, and research papers. I wanted a way to listen instead of read, so I could stay productive while commuting, exercising, or doing chores. Inspired by apps like Speechify and NaturalReader, I aimed to create a lightweight, multilingual, and more affordable alternative that anyone could use.

What I Learned

Through this project, I deepened my understanding of:

• Text-to-Speech (TTS) models and how voice synthesis can feel more natural with prosody and context awareness.
  
• Cross-platform Flutter development, ensuring smooth performance on both iOS and Android.
  
• Subscription infrastructure with RevenueCat, including entitlements, product configurations, and paywall optimization.
  
• The business side of AI apps: ASO (App Store Optimization), localization into 80+ regions, and how monetization strategies vary across countries.
  

How I Built It

The project combines several technologies and services:

• Frontend: Flutter (Dart) with clean architecture, Slang for compile-time i18n, and custom theming.
  
• Backend: Firebase (Firestore, Functions, Auth) + Cloudflare R2 for audio storage & delivery.
  
• AI Voices: Integration with 3rd-party TTS providers and open-source models for high-quality audio.
  
• Monetization: RevenueCat powers the subscription and entitlement system, enabling flexibility with free tiers, premium plans, and token-based virtual currency.
  

Built With

• amazon-web-services
• auth
• azure
• cloudflare
• cloudflare-r2
• dart
• drift-(sqlite)
• firebase
• firebase-(firestore
• flutter
• functions)
• python
• revenuecat
• supabase