Ovi: A Menstrual Health & Wellness App

Problem Statement

Current period trackers often act as passive logs rather than active health companions. There is a significant gap in offering actionable, personalized wellness support and spreading awareness about reproductive/gynecological disorders. Users need a tool that doesn't just track data but helps them truly understand how their menstrual cycle affects their body.

Our Solution

Ovi is a symptom-first menstrual health app designed to identify patterns associated with hormonal risks like PCOS and PMDD. By combining machine learning diagnostics with interactive, adaptive coping strategies (like AI-powered yoga), Ovi empowers women to take control of their reproductive health.

Core Features

Hormonal Risk Screening: An integrated XGBoost Classifier analyzes user data to flag potential indicators of PCOS (with intention to add other disorders like PMDD and Endometriosis), providing users with educational resources for next steps.

AI Yoga Coach: A built-in, real-time wellness assistant. Using your camera feed, the app tracks your skeleton joints to provide accuracy metrics and guide you through poses specifically beneficial for menstrual health.

Smart Period & Symptom Tracking: A comprehensive calendar interface to log flow intensity and physical/emotional symptoms to visualize your cycle patterns.

Key Links

Yoga Coach Demo

Ovi.Yoga.Coach.Demo.10.MB.mp4

Tech Stack

graph TD
    %% Styling Definitions based on your screenshot
    classDef berry fill:#B8336A,stroke:#7a2246,stroke-width:2px,color:#fff
    classDef wisteria fill:#C490D1,stroke:#a66fb3,stroke-width:2px,color:#fff
    classDef periwinkle fill:#ACACDE,stroke:#8e8ebf,stroke-width:2px,color:#2e2e4e
    classDef icyBlue fill:#ABDAFC,stroke:#89b4d1,stroke-width:2px,color:#2e2e4e
    classDef azureMist fill:#E5FCFF,stroke:#c4dadd,stroke-width:2px,color:#2e2e4e

    subgraph Frontend_Layer [User Interface]
        A[Streamlit Web App]:::berry --- B[Custom CSS Styling]:::berry
    end

    subgraph Logic_Layer [Processing & AI]
        C[Camera Feed]:::wisteria --> D[MediaPipe: Skeleton w/ 33 Joints, Kaggle Yoga Poses Data Set]:::wisteria
        D --> E[Python Math: Joint Angle Calculations]:::wisteria
        E --> F[Accuracy Metrics: Visual & Numerical]:::wisteria
        
        G[(Kaggle Menstrual Health Data)]:::periwinkle --> H[Feature Engineering]:::periwinkle
        H --> I[XGBoost Classifier for PCOS, PMDD, etc.]:::periwinkle
    end

    subgraph Data_Layer [Data]
        J[(Python DB Storage)]:::icyBlue
    end

    %% Flow Connections
    A <--> J
    F --> A
    I --> A
    J <--> I

    %% Container Styling
    style Frontend_Layer fill:#E5FCFF,stroke:#B8336A,color:#B8336A
    style Logic_Layer fill:#f9f4fb,stroke:#C490D1,color:#C490D1
    style Data_Layer fill:#E5FCFF,stroke:#ABDAFC,color:#ABDAFC

Loading