Your question is Outlier Handling for Claims. 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 systematically identify and handle outliers in a dataset where extreme values might actually represent valid, high-severity insurance claims at AXA XL Insurance?