How to add features of different sizes (credit histories) to the ML model?

Question

Python

There is a classification model for searching for fraudsters in insurance, where one of the factors includes credit scoring, the object is the insured. In addition, there is data on credit histories. I would like to add new features from there. But each person can have a different number of loans. What is the way to take into account all the credit history data (namely, all data on loans, and not counters about their number, etc.)?

Answer 1

Aggregation per person: maximum / minimum / average loan size. The same approach with the rest of the properties (terms, percentages, etc.).

Answer 2

Structure of used loans by types of loans
Structure of monthly payment by types of loans
Monthly payment by types of loans to average monthly income (if any)
Amount of early repayments for the period to the amount of loans used