Empower Health

• 

🌟 Inspiration

Preventive healthcare often feels inaccessible—especially when it comes to chronic conditions like diabetes. Millions live unaware of their risk, either due to limited screening options or lack of health literacy. Inspired by the AITHON: AI for Social Good theme, we envisioned Empower Health as a way to democratize early diabetes risk assessment using AI, making it fast, personalized, and free—right from your browser.

⚙️ What it does

Empower Health is an AI-powered web application that:

Allows users to securely input key health metrics (Glucose, BMI, Age, etc.)

Predicts diabetes risk in real-time using a trained Random Forest classifier

Generates a professional, downloadable PDF report in one of three design templates (Professional, Modern, Minimal)

Provides actionable health recommendations tailored to user input

Fully deployed and publicly accessible, eliminating any need for setup

🧱 How we built it

We built Empower Health using a focused tech stack optimized for rapid deployment and AI integration:

Python + Flask for backend logic and RESTful APIs

Scikit-learn + Joblib for building and serving the machine learning model

ReportLab for dynamic, multi-template PDF generation

Pandas & NumPy for data preprocessing and input validation

Render for seamless live deployment

All components are version-controlled via Git + GitHub, and built collaboratively under a 48-hour hackathon deadline.

🧩 Challenges we ran into

Dependency Conflict during Deployment: Faced a persistent error on Render due to outdated setuptools, which prevented live deployment. After hours of debugging, we resolved it by updating dependencies (setuptools==70.0.0).

Balancing Simplicity and Utility: Designing a user interface that was both beginner-friendly and capable of delivering medical insights required thoughtful planning.

PDF Layout Consistency: Creating multiple dynamic PDF styles without layout distortion took extensive testing with ReportLab.

🏅 Accomplishments that we're proud of

✅ Successfully deploying the project live within the hackathon window

✅ Integrating dynamic, personalized PDF report generation—rare in short sprint hackathons

✅ Achieving a smooth, end-to-end workflow from form input to prediction to report

✅ Staying aligned with ethical AI principles by focusing on clarity, user empowerment, and interpretability

📚 What we learned

How to build a functional AI prototype under real-world time constraints

The importance of user-first design, especially in health-focused applications

Gained hands-on experience in deployment debugging, particularly with platform-specific issues

Realized the value of storytelling, UI/UX, and accessibility in maximizing the social impact of an AI solution

Built With

• flask
• joblib
• machine-learning
• numpy
• pandas
• python
• render
• reportlab
• scikit-learn