I just did Principal Component Analysis, Random Forest modeling and KMeans Clustering of client data using binary data points of over 700M records.