🌟 Project Story: MoodMirror AI
💡 Inspiration
In a world where mental wellness often takes a backseat, I wanted to create something simple yet impactful — a digital companion that encourages self-expression and emotional reflection. MoodMirror AI was inspired by the idea that sometimes, just being heard — even by an AI — can make a big difference. The rise of voice tech and emotional AI made it possible to combine both for personal wellbeing.
🛠️ How I Built It
MoodMirror was built using:
Bolt.new for the overall web app framework and rapid deployment.
ElevenLabs API to convert user reflections into calming voice responses.
Supabase to manage backend storage and potential user data scaling.
Custom CSS & UI inspired by calm, gradient-based color palettes to reflect the emotional theme.
Users type how they feel, and the app responds with a voice message that mirrors their mood in a comforting tone — like talking to a non-judgmental friend.
📚 What I Learned
How to integrate ElevenLabs voice AI effectively into a live app.
How to use Supabase to prepare a project for large-scale growth and real-time interactions.
The importance of UX design when dealing with mental health-related apps.
How small interactions like voice feedback can increase emotional connection in a product.
⚙️ Challenges Faced
Getting SSO and login flow to work smoothly with Supabase and Entri was tricky.
Fine-tuning the voice output so it sounds natural and comforting took several iterations.
Ensuring the app remained lightweight but powerful, especially on mobile, required testing and optimization.
🚀 What's Next
Adding sentiment analysis and dynamic voice modulation based on emotion.
Introducing multi-language support for broader accessibility.
Publishing on a custom domain to fully meet the hackathon challenge requirements.
Built With
- bolt.new
- elevenlabapi
- javascript
- netlify
- openai
- react.js
- supabase
- tailwind

Log in or sign up for Devpost to join the conversation.