🧠 Inspiration
Mental wellness is often overlooked until it becomes a crisis. Existing apps either feel overly clinical, collect excessive personal data, or reduce emotional wellbeing to simple mood tracking.
We wanted to build something different—a calm, thoughtful space where people could pause, reflect on their thoughts, better understand their emotions, and develop healthier daily habits without sacrificing privacy.
Eunoia was inspired by the idea that technology should help people understand themselves, not replace professional mental healthcare. Instead of diagnosing users, the platform encourages emotional awareness, self-reflection, and positive behavioral change through responsible AI.
💡 What it does
Eunoia is an AI-powered emotional wellness platform that helps users privately reflect on their emotions through natural journaling.
Users can write about their day and receive:
- Emotional analysis
- Wellbeing score
- Mood detection
- Emotional tone identification
- Supportive insights
- CBT-inspired cognitive reframing
- Reflection prompts
- Wellness suggestions
- Emotional trend tracking
The platform also includes breathing exercises, grounding activities, emotional history, downloadable reports, and crisis support resources.
Everything is designed around a privacy-first, local-first philosophy.
🏗️ How we built it
Eunoia was built using a modern React architecture focused on performance, maintainability, and accessibility.
Frontend
- React 19
- TypeScript
- Vite
- Tailwind CSS
- TanStack Router
- React Query
- Radix UI
AI
The emotional analysis system combines:
- Hugging Face Router
- Structured AI outputs
- Function calling
- Rule-based fallback scoring
This ensures users continue receiving meaningful emotional feedback even if AI responses fail.
Privacy
User wellness history remains local to the device.
Security features include:
- Local passcode protection
- SHA-256 hashed PIN storage
- Local-first emotional history
- Downloadable exports
⚡ Challenges we ran into
Designing AI for emotional wellness required balancing usefulness with responsibility.
Some of our biggest challenges included:
- Preventing AI from sounding overly clinical or diagnostic.
- Making emotional feedback supportive without creating dependency.
- Handling unreliable AI responses using fallback analysis.
- Designing a calming user experience that feels safe and approachable.
- Protecting user privacy while still providing personalized experiences.
🏆 Accomplishments that we're proud of
- Built a complete AI-powered emotional wellness platform.
- Designed a clean and calming user experience.
- Created a privacy-first architecture.
- Implemented structured AI with reliable fallback logic.
- Added emotional trend tracking and cognitive reframing tools.
- Open-sourced the project for the community.
📚 What we learned
Building AI for wellbeing goes far beyond integrating an LLM.
We learned the importance of responsible AI design, thoughtful UX, graceful fallbacks, accessibility, and privacy-first engineering. Small design decisions can significantly impact how supported and comfortable users feel while interacting with AI.
🚀 What's next
Our vision is to continue evolving Eunoia into a more personalized emotional wellness companion.
Future plans include:
- Personalized wellness recommendations
- Better long-term emotional insights
- Improved AI memory with user control
- Guided meditation sessions
- Wearable integrations
- Therapist resource integrations
- Enhanced accessibility features
Built With
- ai
- css
- html
- huggingface
- javascript
- radixui
- react
- reactquery
- rechart
- tailwindcss
- tanstackrouter
- typescript
- vercel
- vite
- zod
Log in or sign up for Devpost to join the conversation.