Q52
Single choice
An insurance company developed a new experimental machine learning (ML) model to replace an existing model that is in production. The company must validate the quality of predictions from the new experimental model in a production environment before the company uses the new experimental model to
serve general user requests.
Which one model can serve user requests at a time.
The company must measure the performance of the new experimental model without affecting the current live traffic Which solution will meet these requirements?