Your question is Fraud Class Imbalance Handling. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How do you handle severe class imbalance in a dataset used for fraud detection at GEICO?
Explain a practical modeling and evaluation approach. Address data splitting, resampling or cost-sensitive learning, threshold selection, suitable metrics, calibration, and how you would prevent leakage and monitor the model after deployment.