Implementing Pyvorin for MLOps

Model Serving

Compile preprocessing and postprocessing around model inference.

def predict(input_data):
    features = compiled_preprocess(input_data)
    return model.predict(features)

Monitoring

Compile drift detection and data quality metrics.

Pipeline Orchestration

Compile Kubeflow and MLflow pipeline step logic.