A cross-platform app that lets women log cycles, symptoms and lifestyle data in one place. A machine-learning model then gives an early PCOS risk indication to discuss with a doctor.
The challenge
01- Many women with PCOS go undiagnosed for years.
- Symptoms like irregular cycles, weight change, acne and mood swings show up at different times. Most period trackers never link them to PCOS.
- The client wanted one app to log everything, show patterns over time and give an early risk signal.
The approach
02- A calm, premium pink-and-dark design, with one-tap logging from the home screen.
- Two ML models were trained and compared on a public clinical PCOS dataset, and the stronger one went into production.
- Insights update automatically as users log, and every result comes with plain-language recommendations.
What we built.
Cycle tracking
Periods, flow intensity, automatic cycle length and regularity scoring.
Daily logging
10+ PCOS symptoms with severity, plus mood, exercise, meals with calories, and medication reminders.
Insights dashboard
Charts for symptom trends, mood distribution and cycle analysis.
ML risk assessment
Risk score, key contributing factors, personalised recommendations and history.
Health reports
Reports built from the user's own data, ready to share with a doctor.
Journal & community
A private journal and a supportive community space.
Calendar & reminders
Calendar view with notifications for periods and medication.
Secure by default
Full dark mode and JWT-based authentication.
The results.
A working cross-platform app with a Django REST backend and a validated ML screening model. It turns scattered symptoms into one clear picture users can take to their doctor.
- * The PCOS assessment is a screening aid, not a medical diagnosis.
- * Accuracy comes from cross-validation on a public dataset. Screenshots use demo data.
Inside the product.
Click any screen to enlarge
