CalmMe

🌱 Inspiration

In today’s fast-paced world, mental wellness often takes a backseat. I wanted to create something accessible, empathetic, and intelligent — a safe space where users can check in with themselves, receive emotional support, and get personalized wellness suggestions.

Inspired by apps like Woebot and Wysa, I envisioned CalmMe as an AI-powered friend — someone who listens without judgment and helps users navigate stress, anxiety, and emotional overload. Thanks to tools like Hugging Face, Groq, and Gradio, I was able to bring this idea to life.

💡 What it does

CalmMe is a mental wellness companion that:

• Chats with users empathetically to support emotional well-being
  
• Offers calming suggestions, mindfulness prompts, and affirmations
  
• Understands spoken input (via Whisper STT) and can speak back (via Google TTS)
  
• Remembers context during sessions to maintain a natural flow
  
• Runs entirely on Hugging Face Spaces using powerful language models
  

🛠️ How we built it

• Frontend: Built using Gradio for a clean and interactive UI
  
• AI Chat: Powered by Groq’s LLaMA-3 for fast, intelligent dialogue
  
• Memory & RAG: Managed via LangChain and Chroma DB
  
• Voice Features:
  
  • Whisper for converting speech to text
    
  • Google Cloud TTS for generating soothing voice responses
    
• Data Storage: Used Supabase for anonymous feedback logging and minimal analytics
  
• Deployment: Hosted on Hugging Face Spaces for accessibility and ease of use
  

🧗 Challenges we ran into

• Maintaining context in long conversations without memory drift
  
• Tuning the tone to feel empathetic yet not robotic or vague
  
• Managing API latency and model loading times before integrating Groq
  
• Balancing features with the resource limits of free hosting
  
• Ensuring privacy and anonymity without compromising on interactivity
  

🏆 Accomplishments that we're proud of

• Created a fully functional mental health AI that feels human and calming
  
• Integrated multiple advanced technologies seamlessly (Groq, Whisper, TTS, Supabase)
  
• Designed an experience that is non-judgmental, private, and user-friendly
  
• Built and deployed the entire system solo, from concept to production
  

📚 What we learned

• How to combine LLMs with speech tech, vector DBs, and Gradio effectively
  
• Best practices in building emotionally intelligent AI conversations
  
• Realized the potential of Hugging Face + Groq for real-time mental wellness apps
  
• Learned how important UX is in sensitive domains like mental health
  

🚀 What's next for CalmMe

• Add a daily mood tracker with beautiful graphs and emotional journaling
  
• Introduce personalized breathing exercises and guided meditations
  
• Support multilingual conversation for global accessibility
  
• Release the CalmMe Mobile App for Android and iOS
  
• Open-source the project and invite contributors
  

Built With

• gradio
• groq
• langchain
• python
• rag
• vectordb
• whisper