A FastAPI-based machine learning service that predicts student assessment scores based on their learning characteristics. The API uses a trained ML model to predict student performance based on comprehension, attention, focus, retention, and engagement time metrics.
API Base URL: https://student-backend-2919.onrender.com
API Documentation: https://student-backend-2919.onrender.com/docs
- Machine Learning Predictions: Predict student assessment scores using trained ML models
- Batch Processing: Support for predicting multiple students (up to 200) in a single request
- Authentication & Authorization: JWT-based authentication with API key validation
- Caching: Redis-based caching for improved performance
- Monitoring: Prometheus metrics integration for monitoring API performance
- Logging: Comprehensive logging middleware for request tracking
- Containerized: Docker and Docker Compose support for easy deployment
- Framework: FastAPI 0.110.0
- ML Libraries: scikit-learn 1.3.2, pandas 2.0.3, numpy 1.26.4
- Authentication: python-jose 3.3.0 (JWT tokens)
- Caching: Redis 5.0.4
- Monitoring: Prometheus + Grafana
- Deployment: Docker, Docker Compose
- Server: Uvicorn
POST /login- Authenticate and get JWT token
POST /predict- Predict student score (single or batch)
/metrics- Prometheus metrics endpoint
- Python 3.10+
- Docker & Docker Compose (optional)
- Redis (if running locally)
-
Clone the repository
git clone <repository-url> cd student-backend
-
Install dependencies
pip install -r requirements.txt
-
Set environment variables
# Create .env file API_KEY=your-api-key JWT_SECRET_KEY=your-secret-key REDIS_URL=redis://localhost:6379 -
Start Redis (if running locally)
redis-server
-
Run the application
uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload
-
Using Docker Compose (Recommended)
docker-compose up -d
This will start:
- FastAPI application on port 8000
- Redis on port 6379
- Prometheus on port 9090
- Grafana on port 3000
-
Using Docker only
docker build -t student-backend . docker run -p 8000:8000 student-backend
First, get an authentication token:
curl -X POST "https://student-backend-2919.onrender.com/login" \
-H "Content-Type: application/json" \
-d '{
"username": "admin",
"password": "admin"
}'Response:
{
"access_token": "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9..."
}curl -X POST "https://student-backend-2919.onrender.com/predict" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_TOKEN" \
-H "X-API-Key: demo-key" \
-d '{
"comprehension": 75.5,
"attention": 82.3,
"focus": 78.9,
"retention": 80.1,
"engagement_time": 95
}'Response:
{
"predicted_score": "78.45"
}curl -X POST "https://student-backend-2919.onrender.com/predict" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_TOKEN" \
-H "X-API-Key: demo-key" \
-d '{
"students": [
{
"comprehension": 75.5,
"attention": 82.3,
"focus": 78.9,
"retention": 80.1,
"engagement_time": 95
},
{
"comprehension": 68.2,
"attention": 74.8,
"focus": 71.3,
"retention": 72.5,
"engagement_time": 87
}
]
}'Response:
{
"batch_size": 2,
"predictions": [
{"predicted_score": "78.45"},
{"predicted_score": "71.23"}
]
}The API expects the following student characteristics:
| Parameter | Type | Description | Range |
|---|---|---|---|
comprehension |
float | Student's comprehension level | 0-100 |
attention |
float | Student's attention span | 0-100 |
focus |
float | Student's focus ability | 0-100 |
retention |
float | Student's retention capacity | 0-100 |
engagement_time |
integer | Time spent engaged (minutes) | 0+ |
- JWT Authentication: All prediction endpoints require valid JWT tokens
- API Key Validation: Additional API key validation for enhanced security
- Rate Limiting: Built-in protection against abuse
- Input Validation: Comprehensive input validation using Pydantic models
Access metrics at: https://student-backend-2919.onrender.com/metrics
- Local: http://localhost:3000 (when using Docker Compose)
- Default credentials: admin/admin
The application includes comprehensive logging middleware that tracks:
- Request/response times
- API usage patterns
- Error rates
- Cache hit/miss ratios
student-backend/
├── app/
│ ├── api/
│ │ ├── routes_auth.py # Authentication endpoints
│ │ └── routes_predict.py # Prediction endpoints
│ ├── cache/
│ │ └── redis_cache.py # Redis caching logic
│ ├── core/
│ │ ├── config.py # Application configuration
│ │ ├── dependencies.py # Dependency injection
│ │ ├── exceptions.py # Exception handlers
│ │ └── security.py # JWT and security utilities
│ ├── middleware/
│ │ └── logging_middleware.py # Request logging
│ ├── models/
│ │ └── model.joblib # Trained ML model
│ ├── services/
│ │ └── model_service.py # ML prediction logic
│ └── main.py # FastAPI application entry point
├── data/
│ ├── student_performance.csv # Raw data
│ └── students_cleaned.csv # Processed data
├── notebooks/
│ ├── analysis.ipynb # Data analysis
│ ├── assignment_score.ipynb # Score analysis
│ └── synthetic_data.ipynb # Data generation
├── training/
│ ├── train_model.py # Model training script
│ └── train_utils.py # Training utilities
├── docker-compose.yml # Multi-service setup
├── Dockerfile # Container definition
├── requirements.txt # Python dependencies
└── prometheus.yml # Prometheus configuration
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add some amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
This project is licensed under the MIT License - see the LICENSE file for details.
For support and questions:
- Create an issue in the repository
- Check the API documentation at
/docsendpoint - Review the monitoring dashboards for system health
- v1.0.0 - Initial release with basic prediction functionality
- v1.1.0 - Added batch prediction support
- v1.2.0 - Implemented caching and monitoring
- v1.3.0 - Enhanced security and authentication
Built with ❤️ using FastAPI and scikit-learn