Automated Feature Extraction (AFE) enables the creation of automated and supervised extraction of an optimal set of predictive features from transactional datasets without human effort. It helps to transform temporal and relational datasets into feature matrices for modeling. The user just needs to fill in a few configurations about the raw data source. Then, AFE generates a lake of input features from transactional tables. Although the input lake may be large at the level of millions of variables, AFE intelligently benefits from supervised and unsupervised methods to reduce the scale of the lake to the thousands. The fact that AFE generates features directly from transactional tables without any data preprocessing is a feature that other AutoML products do not have. AFE makes Autonon AutoML a unique solution created by Organon Analytics.