Who will be churned in the next 4 months?

The task is predicting churn for a given time horizon (for example, 4 months or 6 months in the future). The standard approach predicts only that somebody will churn or not. Is there any approach that can solve this problem? Is it necessary to organize features on time basis in order to predict churn in the next period?

I have found this short explanation: https://stackoverflow.com/questions/64237069/predicting-customer-churn-over-a-period-of-time

but it is not clear, is there connection between prediction period and sliding window? How to define features in one window, and what is the best period for sliding? Is the sliding window only approach for this problem.

Could somebody provide example how to define features for sliding window or generally how to solve this problem?