A company has just released a new machine learning model (B) and wants to test it using the A/B testing method. A Sagemaker’s endpoint is already in service, and a model (A) is already serving the traffic, however, the company believes that the newer model will serve better as it had outstanding results in the training set. The two models classify a product’s image as faulty or not. How should the company update the Sagemaker’s endpoint?