🧠What I Built
I constructed Smile2Earn, a web application that employs AI-powered smile recognition to reward users with virtual money in the form of SMILE Coins whenever they smile. It was our aim to create something significant and happy — a system that converts positivity into tangible value.
- Users authenticate with Supabase
- Smile recognition is performed in-browser with TensorFlow\.js
- Users receive SMILE Coins (100 SMILE = â‚ą1)
- A wallet keeps coins earned
- Withdrawal system (through UPI or bank) is developed but on hold
- Revenue is on a plan through ad monetization
⚙️ Technologies Utilized
- Bolt.AI – for boilerplate generation, bug fixes, and rapid logic writing
- TypeScript + Tailwind CSS – for responsive and clean UI
- TensorFlow\.js – for smile detection in real-time
- Supabase – for authentication and database storage
- Vite – rapid build setup
- Secure Storage – for session data and tokens
⚒️ Challenges I Overcame
- Conflicts of Dependencies
- Dependencies such as
@nativescript/firebaseversions were outdated or non-existent - Bolt assisted in producing alternative code with improved support
- Smile Detection Logic
- Timing TensorFlow's smile detection to prevent false positives
- Detecting in a privacy-safe and light manner for real-time application
- Authentication Flow
- Merging Supabase's auth into my own custom frontend logic
- Bolt assisted in creating working login, signup, and logout handlers
⚡ How Bolt AI Altered My Workflow
- Time saved correcting errors, maintaining code versions, and sorting out npm problems
- Created fully functional Supabase integration
- Allowed me to test concepts quicker without being caught in technical dead-ends
- Played the role of coding mentor when I encountered roadblocks
📽️ Watch the Demo
👉 Smile2Earn – Earn Money by Smiling (YouTube Video)
🎯 Final Thoughts
This wasn't just about code — it was about creating something joyful.
Thanks to Bolt.AI, I wrote an app that pays individuals for smiling — and that, to my mind, is a huge success.
Top comments (0)