You have trained several candidate models on the same labeled dataset, and the validation scores are close enough that you are not confident which one will generalize best. The team wants a clear way to compare them before picking a final model.
What is cross-validation, and why is it important in model selection?
