This project demonstrates the successful deployment of a Machine Learning classification model using a RESTful API developed with FastAPI. The service predicts the credit risk (Good/Bad) of a customer based on 48 processed features derived from the German Credit Data.
This project showcases competence in critical MLOps (Machine Learning Operations) and deployment concepts:
- API Framework: FastAPI (Chosen for high performance and automatic documentation).
- Server: Uvicorn (ASGI server used for running the FastAPI application).
- Model: Logistic Regression (Trained using Scikit-learn).
- Model Serialization: Joblib (Used to serialize/save the trained model: german_credit_model.pkl).
- MLOps Focus: Model Deployment, API Development, and Dependency Management (requirements.txt).
To test the deployment locally, follow these steps:
git clone [https://github.com/soheilon21-a11y/German_Credit_Risk_API_Deployment.git](https://github.com/soheilon21-a11y/German_Credit_Risk_API_Deployment.git)
cd German_Credit_Risk_API_Deployment
2. Install Dependencies
All required packages (FastAPI, Uvicorn, Pandas, etc.) are listed in requirements.txt.
pip install -r requirements.txt
3. Start the Server
Run the Uvicorn server from the project directory:
python -m uvicorn main:app --reload
(The server will start running on http://127.0.0.1:8000)
🧪 How to Test the Prediction Endpoint
You can interact with the deployed model using the automatic documentation interface (Swagger UI):
Access Docs: Open your web browser and navigate to: http://127.0.0.1:8000/docs
Perform Prediction:
Click on the /predict_risk endpoint.
Click Try it out.
Paste the following sample of 48 processed features into the Request body:
{
"features": [1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
}
Click Execute to receive the prediction result (e.g., "Good Credit Risk").