MoodMesh – AI Therapist & Fitness Coach

Ai powerd platform that understands your health patterns by analyzing your health through mood logs and conversations to suggest tailored advice. It also helps with physical exercise as a trainer.

Comment

Inspiration

Mental health care is often reactive, expensive, and inaccessible when people need it most. We wanted to change that. The idea for MoodMesh came from a simple thought what if emotional understanding could live in your pocket, 24/7? Something that listens without judgment, offers real help, and learns your emotional rhythm over time.

We combined our passion for AI, psychology, and design to build a platform that acts as a personal therapist, fitness coach, and support community all working together to help users build healthier minds and habits.

What it does

MoodMesh is an AI-powered mental health and wellness companion that offers:

AI Therapist: A Gemini-powered conversational therapist that uses CBT, DBT, and mindfulness techniques to provide evidence-based emotional support. Mood Analytics: Smart mood logging and visualization to track emotional trends, triggers, and progress over time. Crisis Detection: Real-time analysis of user input to identify signs of emotional distress and alert with emergency resources. Community Support: Safe, moderated peer spaces for people to connect, share, and heal together. Fitness & Meditation Coach: Personalized exercise and breathing recommendations based on emotional state. Gamified Motivation: Achievements, badges, and wellness stars that keep users motivated through positive reinforcement.

How we built it

We built MoodMesh as a full-stack AI platform combining modern web technologies with advanced AI capabilities:

Frontend: React 19 + TailwindCSS for a responsive, calming UI Backend: FastAPI (Python) for performance and async handling AI Engine: Google Gemini 2.5 Flash for natural, context-aware conversations Database: MongoDB with Motor async driver for mood logs and user data Authentication: Secure JWT system with bcrypt password hashing Realtime Communication: Socket.IO for live chat in communities DevOps: Dockerized setup for easy deployment and scaling Data Privacy: Encrypted user data and strict isolation between mental health logs and community chats

Challenges we ran into

Integrating real-time emotion detection and crisis keyword analysis without false positives. Designing a tone of empathy for the AI that feels supportive, not robotic. Managing CORS and WebSocket connections between FastAPI and React during deployment. Achieving low latency in conversations powered by large AI models. Ensuring mental health features remain safe, non-diagnostic, and privacy-compliant.

Accomplishments that we're proud of

Built a working AI therapist that adapts to users’ moods in real time. Integrated crisis detection and safety plans to offer tangible mental health support. Designed a gamified mental wellness system that makes emotional tracking rewarding. Developed a calming, accessible UI using soft color palettes and clean typography. Created a functioning AI + community hybrid that merges technology with human empathy.

What we learned

How to fine-tune AI communication style to balance empathy and professionalism. The importance of ethical design in mental health AI transparency, safety, and user control matter. Deep understanding of real-time data flow between React and FastAPI using Socket.IO. How gamification can positively influence mental wellness engagement. That small UI touches tone, animations, and whitespace have a big impact on user comfort.

What's next for MoodMesh - AI Therapist & Fitness Coach

Voice Tone Analysis: Integrate audio-based mood detection. Mobile App: Build cross-platform React Native version for iOS and Android. Professional Marketplace: Allow verified therapists and trainers to join the ecosystem. Multilingual Support: Add 10+ languages to make emotional support globally accessible. Wearable Integration: Connect with smartwatches for real-time stress and sleep tracking. Advanced ML Personalization: Evolve MoodMesh into an adaptive, emotionally intelligent coach that truly understands the user over time.

Built With

Share this project:

Updates