A mobile and wearable health-tracking platform designed to monitor biometric vitals in real time and detect potentially risky health conditions. The system consists of a Swift-based iOS application, an Apple Watch application, and a backend API. Users can monitor their health status, receive alerts when abnormal patterns are detected, and access relevant medical professionals based on their condition and location.

The objective of this project is to develop a real-time health monitoring platform that centralizes biometric data collection, risk analysis, and medical response. The system enhances user safety by detecting abnormal health signals early, providing clear risk explanations, and enabling fast access to appropriate local physicians through intelligent matching.

Biometric Data Collection

• Collect health vitals such as heart rate and body temperature from the Apple Watch.
• Transmit biometric data at short, regular intervals for real-time monitoring.

Health Monitoring

• Display current health status within the iOS application.
• Provide a native Swift-based interface optimized for iOS and watchOS.

Alerts & Guidance

• Notify users when vitals exceed predefined risk thresholds.
• Present clear guidance on next steps when a health risk is detected.

Risk Detection

• Analyze incoming biometric data to compute a risk score.
• Classify severity levels (low, medium, high).
• Identify the primary health risk category based on detected patterns.

Contextual Explanation

• Return short explanations describing why a risk was triggered.
• Incorporate country or regional context into risk interpretation.

Medical Specialty Matching

• Map detected health risks to relevant medical specialties.
• Generate short specialist recommendations based on user vitals.

Location-Based Filtering

• Filter recommended doctors by geographic location.
• Prioritize nearby medical professionals for faster response.

Semantic Search

• Use embeddings and MongoDB vector search to match physicians by medical relevance.
• Fall back to specialty-based matching or local datasets when vector search is unavailable.

API Services

• Backend API built with Node.js and Express.
• Dedicated endpoints for health analysis and doctor recommendations.

Database Management

• MongoDB used as the primary database.
• Store physician data, health risk results, and vector embeddings.
• Local JSON fixtures used as fallback data sources.

Security

• Environment variables used for API keys and configuration.
• No sensitive credentials committed to version control.

• Nita Lefebvre
• William Huang
• Brandon Falcon Aguilar
• Rakib Akter

• Mobile & Watch Applications: Swift, SwiftUI, iOS, watchOS
• Backend: Node.js, Express
• Database: MongoDB
• Infrastructure: Docker
• Middle wear: Multer, Axios
• AI & Search: Google Gemini API, MongoDB Atlas Vector Search