MySugarCare

Inspiration

Diabetes is a rapidly growing global health issue, affecting millions every year. Many people are unaware of their risk until symptoms become serious. We wanted to create a tool that empowers individuals to proactively understand their risk using real-world data and personalized insights.

What it does

MySugarCare is a mobile-first app that predicts a user's risk of developing Type 2 Diabetes using a combination of lifestyle, health, and behavioral data. Users input information about their diet, physical activity, stress levels, health conditions (like hypertension or PCOS), and more. The app uses a machine learning model to assess this data and give a risk score, accompanied by tailored suggestions to reduce risk.

How we built it

• Frontend: We used React Native to build a clean, user-friendly mobile app experience that works on both Android and iOS.
• Backend: Node.js and Express.js handle the backend API, which processes user data and sends it to the ML model.
• ML Model: We used a logistic regression model trained on the Pima Indian Diabetes Dataset, tuned for accuracy using cross-validation.)
• Deployment: The app was tested using Expo Go and deployed there.

Challenges we ran into

• Finding high-quality, relevant datasets for diabetes risk factors.
• Cleaning and preprocessing real-world data (e.g., handling missing values, normalizing input).
• Integrating the machine learning model into the mobile flow while ensuring performance and privacy.
• Designing a UI that is informative yet simple for non-technical users.

Accomplishments that we're proud of

• Creating a working app that gives real-time diabetes risk predictions based on lifestyle.
• Building a complete pipeline from frontend to ML backend with data integration.
• Designing an experience that could realistically help users change behavior and understand their health risks.

What we learned

• How to train, evaluate, and integrate ML models into a mobile app.
• How different risk factors (like PCOS, hypertension, or stress) statistically relate to diabetes.
• Importance of UI/UX when dealing with sensitive health information.
• How to collaborate as a team across design, frontend, and ML responsibilities.

What's next for MySugarCare

• Add more risk domains like sleep habits, family history, and geolocation-based diet risks.
• Deploy on both app stores with secure user login and cloud database support.
• Add chatbot integration (using OpenAI or similar) to answer user questions and suggest daily health actions.
• Expand to other health risk models: heart disease, obesity, and hypertension.

Built With

• express.js
• node.js
• python
• react-native
• scikit-learn