CalmSphere
AI-powered mental wellness platform offering personalized journaling, mood tracking, songs recommendation and community support
Motive
I wanted to create a gentle, all-in-one digital sanctuary for mental wellness and spiritual growth. Recognizing the need for a safe, private space, I built Calm Sphere to consolidate mindful journaling, empathetic AI companionship, and mood tracking into a single, healing ecosystem.
Overview
Calm Sphere is an AI-powered mental wellness application designed to provide emotional support. It features 'Calm Bot,' a compassionate AI companion for non-judgmental chat, and a smart journaling system where AI analyzes entries to generate supportive feedback and mood scores. The platform also includes AI-curated music therapy playlists based on the user's emotional state and a comprehensive dashboard for tracking mental well-being trends.
User Flow
User signs up -> Accesses Dashboard with daily quotes -> Writes in Journal or Chats with Calm Bot -> AI analyzes content for sentiment and emotion -> System updates Mood Score and generates personalized Song Recommendations -> User reviews wellness insights and tracks progress.
Future Plans
- Public launch and performance optimizations for faster reloads.
- Make the chatbot universally available to all users.
- Evolve the AI into a 'talkative' companion with voice interaction and video creation capabilities for journaling.
- Launch a dedicated mobile application.
- Implement a 'Sharing Ideas Panel' to allow users to connect with others who have similar thoughts and ideas, similar to Reddit.
- Learned how to work with data APIs effectively.
- Integrated multiple AI/LLM models into the application workflow.
- Implemented real-time YouTube data extraction.
- Built context-saving mechanisms for conversational chat applications.
- Explored and implemented animations using Framer Motion.
- Fetching reliable and real-time YouTube data was challenging due to API limits and data consistency issues.
- Identifying and selecting the most appropriate AI/LLM models for each feature and dashboard section required extensive experimentation and evaluation.