Inspiration
University stress often goes unnoticed and unspoken. We wanted to build something that doesn’t just listen, but actually recognises when someone is struggling and responds with care.
What it does
SerenAIty chats with the user, detects emotional tone, and tracks stress patterns over time. When stress increases, it gently offers: Guided breathing exercises, and if stress persists, Suggestions for calming nearby places to help reset.
How we built it
Flask backend for routing and chat logic OpenAI model for emotional tone detection and supportive responses JavaScript frontend for a simple, calm chat interface Lightweight mood log stored in JSON to recognise stress trends
Challenges we ran into
Designing emotional support that feels genuine instead of automated Balancing agentic intervention without being intrusive Keeping the interface simple, calm, and reassuring
Accomplishments that we're proud of
The conversation feels supportive, not just “AI replying” The stress-based intervention flow works smoothly in real use The experience is genuinely calming, which was the goal
What we learned
Emotional logic and timing are just as important as technical logic when designing wellbeing support tools.
What's next for SerenAIty
Voice-guided support using ElevenLabs Smarter location suggestions based on the user’s campus and habits Optional journaling / reflection mode for longer-term wellbeing Debating ambeint background noise / music
Log in or sign up for Devpost to join the conversation.