Nowadays insurance companies offer car usage - based insurance programs designed to leverage data - driven technologies and improve insurance products. This is done by collecting GPS driving data from customers using either a vehicle telematics dongle and/or a smartphone application. Vehicle position and other sensor data is used to characterize driving behavior. Along with other contextual information, these rich data have the potential to provide better customer segmentation, risk estimation and cost prediction. This talk will review the data analytics challenges brought by this massive amount of data in order to derive meaningful driver behavior profiles.