MindPulse

AI-powered mental health companion that analyses how Gen Z students feel through their thoughts and provides personalized wellness insights and coping strategies using Amazon Nova.

Comment

Inspiration

Gen Z is facing a silent mental health crisis. 1 in 3 students report feeling constantly overwhelmed, yet most never seek help due to stigma. I wanted to build something that meets them where they are — no appointments, no judgment, just a safe space to express how they're really feeling.

What it does

MindPulse is an AI-powered mental wellness companion for Gen Z. You type how you're feeling — anything, no structure needed — and Amazon Nova analyzes your text for emotional patterns, stress indicators, and mental health signals. It returns a mood score, detected patterns, personalized coping strategies, an affirmation, and a breathing exercise — all in seconds.

How I built it

  • Frontend: HTML5, CSS3, Vanilla JS — custom dark UI with animations
  • Backend: Python + Flask REST API
  • AI: Amazon Nova Lite via AWS Bedrock — processes natural language and returns structured wellness insights
  • Features: Mood history tracking, Chart.js wellness graph, Box Breathing animation

Challenges I ran into

  • Learning the Amazon Nova API request format (different from other LLMs)
  • Designing a UI that feels safe and calming, not clinical
  • Getting Nova to return consistent structured JSON responses

What I learned

How to integrate Amazon Nova via AWS Bedrock, prompt engineering for mental health contexts, and how to build emotionally sensitive AI products.

What's next

Voice input, multilingual support for regional Indian languages, and a crisis helpline integration for high-urgency detections.

Built With

Share this project:

Updates