Real-time health anomaly detection and AI-powered fitness insights. Built across 4 milestones at Infosys Springboard.
| Feature | Description |
|---|---|
| 🧠 AI Health Insights | Claude AI analyzes your metrics and generates personalized recommendations |
| Z-score statistical analysis flags abnormal heart rate events in real-time | |
| 📈 7-Day Forecasting | Time-series prediction for heart rate, steps, and sleep |
| 📊 Comparative Analytics | Daily aggregation and trend visualization across all metrics |
| 🏆 Wellness Score | 0–100 composite score weighted across heart health, activity, and sleep |
| 🧬 Behavior Profiling | ML classification: Sedentary → Very Active with improvement plans |
| 📥 CSV Upload | Analyze your own wearable data |
| 🌙 GSAP Animations | Cinematic scroll-driven animations and ECG visualizations |
- Next.js 14 (App Router, TypeScript)
- Tailwind CSS — custom design system
- GSAP + ScrollTrigger — cinematic animations
- Framer Motion — micro-interactions
- Recharts — data visualizations
- Anthropic Claude API — health insights generation
- Custom prompting — structured JSON health analysis
- Pandas + NumPy — data processing pipeline
- Streamlit — original ML dashboard
- Prophet — time-series forecasting
- Z-score statistics — anomaly detection
- Vercel — production deployment with edge functions
# 1. Clone the repo
git clone https://github.com/DashamiJituri/FitPulse_Health_Detection_Project.git
cd FitPulse_Health_Detection_Project
# 2. Install dependencies
npm install
# 3. Set up environment
cp .env.example .env.local
# Add your ANTHROPIC_API_KEY from https://console.anthropic.com
# 4. Run locally
npm run dev
# Open http://localhost:3000src/
├── app/
│ ├── page.tsx # Landing page (GSAP animated)
│ ├── dashboard/page.tsx # Main analytics dashboard
│ └── api/health-insights/ # Claude AI endpoint
├── components/
│ ├── landing/ # ECG, StatsCounter, FeatureCard
│ └── dashboard/ # Charts, Gauge, Panels
├── lib/
│ └── healthData.ts # Data generation & ML logic
└── types/
└── health.ts # TypeScript interfaces
Synthetic generation of heart rate, steps, and sleep data. Cleaning, merging, and exporting unified datasets.
Time-series forecasting with Prophet for heart rate, sleep, and step predictions. 7-day horizon with confidence intervals.
Anomaly detection (Z-score), behavior analysis (Sedentary → Very Active), comparative analytics, and wellness scoring.
Integrated all milestones into a production Next.js application with AI insights, beautiful visualizations, and CSV export.
Dashami Govind Jituri 📧 dashamijituri02@gmail.com Built at Infosys Springboard, 2025