🚀 Live Demo: https://predictibetes-ai.onrender.com/
⭐ GitHub Repository: https://github.com/Ashwitha-Ramesh/PredictiBetes-AI
PredictiBetes is a recruiter-ready end-to-end Machine Learning web application that predicts diabetes risk using multiple supervised learning algorithms while providing clinical analytics, explainable AI insights, interactive dashboards, and prediction history management.
Unlike traditional ML notebooks, PredictiBetes demonstrates the complete production workflow—from data preprocessing and model training to deployment with FastAPI and cloud hosting.
- Logistic Regression
- Random Forest
- Decision Tree
- Support Vector Machine (SVM)
- K-Nearest Neighbors (KNN)
- Gaussian Naive Bayes
- Interactive Plotly Dashboard
- Correlation Heatmaps
- Feature Distributions
- Boxplots
- Scatter Plots
- Clinical Statistics
- Dataset Exploration
Each prediction includes
- Risk Probability
- Confidence Score
- Top Positive Risk Factors
- Protective Factors
- Clinical Interpretation
- Personalized Health Recommendations
✔ Single Patient Prediction
✔ Batch CSV Prediction
✔ Real-time Risk Classification
✔ Confidence Estimation
✔ Feature Scaling
✔ Missing Value Handling
- SQLite Database
- Search Predictions
- Filter Records
- Delete Records
- CSV Export
- Glassmorphism Design
- Bootstrap 5
- Responsive Layout
- Interactive Charts
- Dark Theme
- Animated Dashboard
Successfully deployed on
Render
Production Ready
No Docker Required
```
PredictiBetes-AI
│
├── app
│ ├── routers
│ │ ├── predict.py
│ │ ├── analytics.py
│ │ └── history.py
│ │
│ ├── services
│ │ ├── ml_service.py
│ │ └── eda_service.py
│ │
│ ├── static
│ │ ├── css
│ │ ├── js
│ │ └── images
│ │
│ ├── templates
│ │ ├── base.html
│ │ ├── index.html
│ │ ├── predict.html
│ │ ├── dashboard.html
│ │ ├── eda.html
│ │ ├── models.html
│ │ └── history.html
│ │
│ ├── database.py
│ ├── main.py
│ ├── ml_model.py
│ ├── models.py
│ ├── schemas.py
│ ├── train_model.py
│ └── utils.py
│
├── dataset
│ └── diabetes.csv
│
├── model
│ ├── best_model.pkl
│ └── metrics.json
│
├── notebooks
│
├── screenshots
│
├── requirements.txt
├── LICENSE
├── CONTRIBUTING.md
├── CHANGELOG.md
└── README.md
- FastAPI
- Python
- SQLAlchemy
- Uvicorn
- Pydantic
- Scikit-Learn
- Pandas
- NumPy
- Joblib
- HTML5
- Jinja2
- Bootstrap 5
- Vanilla JavaScript
- Plotly.js
SQLite
git clone https://github.com/Ashwitha-Ramesh/PredictiBetes-AI.git
cd PredictiBetes-AIWindows
python -m venv venv
venv\Scripts\activateLinux / macOS
python3 -m venv venv
source venv/bin/activatepip install -r requirements.txtpython app/train_model.pyuvicorn app.main:app --reloadOpen
http://127.0.0.1:8000
https://predictibetes-ai.onrender.com/
Hosted using
- Render
- FastAPI
- Uvicorn
- SQLite
Dataset
↓
Data Cleaning
↓
Feature Engineering
↓
Train/Test Split
↓
Model Training
↓
Model Evaluation
↓
Best Model Selection
↓
FastAPI Backend
↓
Interactive Dashboard
↓
Cloud Deployment
| Algorithm | Evaluated |
|---|---|
| Logistic Regression | ✅ |
| Random Forest | ✅ |
| Decision Tree | ✅ |
| Support Vector Machine | ✅ |
| KNN | ✅ |
| Gaussian Naive Bayes | ✅ |
- User Authentication
- Doctor Dashboard
- PDF Clinical Report
- REST API Versioning
- Docker Support
- PostgreSQL Integration
- SHAP Explainability
- Deep Learning Models
- CI/CD using GitHub Actions
Computer Science Engineering Student
AI • Machine Learning • Data Analytics • Open Source
https://www.linkedin.com/in/ashwitha-ramesh-0123ab315/
🐙 GitHub
https://github.com/Ashwitha-Ramesh
If you found this project useful,
⭐ Star the repository
🍴 Fork the repository
📢 Share it with others
Licensed under the MIT License.
Disclaimer: PredictiBetes is an educational Machine Learning project intended for demonstration, research, and portfolio purposes. It is not a substitute for professional medical diagnosis or treatment.