You have built a binary classification model for a consumer lending product that predicts whether an applicant is likely to default. A client wants to understand what the model is doing, why it makes certain predictions, and whether they can trust it enough to use in decision-making.
How would you explain this machine learning model to the client in a way that is accurate, understandable, and actionable? Describe how you would choose the model, evaluate it, and communicate both its predictions and limitations.