Scenario

You've trained a machine learning model and offline results look promising. Before putting it into production, your team wants a clear validation process that checks whether the model is reliable and fit for real use.

Question

How do you validate a machine learning model before deployment?