Inspiration
Teenagers today face unprecedented academic pressures, social dynamics, digital overload, and mental health challenges—often without adequate emotional support. Inspired by the need for a private, always-available, and culturally inclusive tool for emotional expression, Teen Wellness AI (Mendify) was born. The goal was to create a safe space where teens could vent, reflect, and heal—without judgment, fear, or stigma
What it does
Mendify is an AI-powered mental wellness companion designed specifically for teens. Key features include:
1)Voice and image journaling with Whisper-based transcription
2)Anonymous, nickname-based logins (no personal info needed)
3)Multilingual, empathetic AI responses for emotional check-ins
4)Real-time mood insights via multimodal sentiment analysis
5)Personalized micro-habit nudges (hydrate, gratitude, breathwork)
6)Breathing, grounding, and meditation exercises auto-recommended
7)Vision Board Generator
8)AI Home-Work Solver
9)Mood Art Generator
10)AI Chat Bots
Cross-device use via hands-free voice input and whisper-quiet audio feedback
How we built it
Frontend: React.js for modular, responsive UI
Backend: Node.js + Express with MongoDB
Voice-to-text: Integrated OpenAI’s Whisper model locally
AI Responses: Groq LLM backend for instant, contextual replies
Security: End-to-end encryption and anonymous identity
Sentiment Analysis: Text + image mood tracking for trends
Languages: Auto-detect & respond in user’s spoken language
Challenges we ran into
Voice-to-Text Conversion with Whisper As a beginner in AI, I struggled with accurate transcription in varied teen environments—like buses, hallways, and homes.
Google Speech-to-Text failed in noisy conditions and with diverse accents.
Many APIs were slow or privacy-compromising (cloud-based).
How we overcame it
We switched to OpenAI’s Whisper, which:
Handled noisy environments and accent diversity well
Allowed on-device processing, enhancing privacy and speed
Supported multilingual transcription out-of-the-box By integrating Whisper, we significantly improved accuracy, performance, and user trust—turning journaling into a seamless, voice-first experience.
Accomplishments that we're proud of
Achieved real-time, low-latency AI interactions with Groq LLM
Delivered judgment-free journaling via voice, image, or text
Created instant mood feedback from multi-modal inputs
Ensured privacy-first architecture with anonymous access
Developed a UX that feels as easy as texting a friend
What we learned
Power of multimodal AI to understand emotional context
Whisper's capability for real-world voice transcription
How to architect a low-latency, privacy-first journaling tool
Importance of cultural and language inclusivity in mental health tech
Designing for empathy, not just functionality
What's next for Mendify
Voice-synthesized responses for full audio conversations
Gamified mental health challenges to boost resilience
Crisis escalation protocols for detecting at-risk teens
AR mindfulness overlays for immersive grounding
Mobile-first version for use in school corridors and homes
Expand into school wellness programs (B2B partnerships)
Log in or sign up for Devpost to join the conversation.