Blood Donor ML Prediction Website

I have a cleaned dataset containing donor health information and I want a lightweight web app that predicts the likelihood of a person making a future donation. When a visitor submits their details, the model should output a single prediction score on screen—no extra charts or recommendations—then quietly trigger an automatic email with the same score to the user and a copy to me for tracking.

Here is the flow I’m after:
1. Train a supervised ML model on the provided health-related features. Accuracy is important, but keep the pipeline transparent so I can retrain it later if we add more data.
2. Wrap the model in a simple website (Python Flask or Django, React, or any stack you are comfortable with). The form collects the required health inputs, calls the model, and prints only the prediction score.
3. Implement email automation (SMTP or a trusted API such as SendGrid) that fires immediately after each prediction.

Deliverables:
• Trained model file and reproducible training script
• Source code for the web app with clear setup instructions
• Brief README explaining how to retrain and change email credentials

If you have prior work with scikit-learn, TensorFlow, Flask, Django, or similar tools, please mention it. I look forward to seeing a working demo deployed on a free tier (Heroku, Render, or comparable) before hand-off.