To develop a mobile-first yoga therapy app that integrates biometric and behavioral health data to generate personalized wellness plans using predictive analytics. The project aims to support both personal and population-level health improvements by leveraging wearable data, AI-driven personalization, and therapist-guided care.

This capstone will demonstrate the application of data aggregation, mining, visualization, and predictive modeling to optimize chronic condition management, stress reduction, and overall wellness through digital yoga therapy.

Project Outcomes and Deliverables

A. Data Architecture & Aggregation

• Aggregation of biometric data from Apple Health, Garmin, Fitbit, and user-reported inputs (e.g., sleep, mood, pain, HRV, stress levels)
• Integration schema for therapist-generated inputs, session adherence, and journaling metrics

B. Predictive Computational Modeling

• Build and train machine learning models (e.g., logistic regression, random forest, neural nets) to:
• Predict optimal yoga practices based on health and biometric inputs
• Recommend flow adjustments for chronic pain, fatigue, PCOS, insomnia, and more
• Validation metrics: accuracy, F1 score, AUC-ROC

C. Data Mining & Segmentation

• Identify behavioral and physiological usage patterns
• Cluster user personas based on biometric trends and engagement
• Analyze correlations between routine adherence and symptom improvement

D. Visualization Techniques

• Interactive dashboards for:
• Individual clients (personal health insights)
• Yoga therapists (session logs, biometric summaries)
• Administrators (population trends and cohort health progress)
• Built using PowerBI, Looker Studio, or Tableau