How would you handle a scenario where a machine learning model faces data drift in a production system at El Corte Inglés?
