Designing relevant services and superior customer care requires a deep understanding of customer segments, including their demographics. Demographic data about customers is at times incomplete. The purpose of this study is to improve customer demographic data while preserving customer privacy. Our paper discusses a methodology that uses anonymized and aggregated mobility network data combined with US Census data to produce insightful demographic profiles. We estimate an anonymized device user’s demographics by first inferring a home base station based on its mobility network traffic patterns, then associating it with a US Census Block and its corresponding demographics from the American Community Survey. Additional statistical inference is called for when the device home base station can’t be determined with sufficient accuracy. Applying big data technology and scalable data science in a cloud environment, we analyzed and aggregated billions of anonymized mobility network traffic records and customer billing ZIP to develop models that link mobile devices to the most likely demographic profile. The distributional nature of these anonymized and aggregated profiles protects customer privacy. We will discuss the methodology and complexity of constructing the models.