At our archaeological fieldwork project in the South Caucasus country of Armenia, we are capturing extremely large amounts of digital 3D data about both spaces and objects. We built a custom 3D object scanner, with which we have been able to create highly accurate models of tens of thousands of individual pottery sherds. We also use a commercial hand-held scanner to capture models of special objects. At our archaeological site, we create precise 3D models of the complete volume of each unique context that we have excavated in the trench. These models will enable us to recreate a 3Dpuzzle of exactly interlocking contexts from our entire excavation, providing the opportunity to carefully study all stratigraphic relationships. After briefly presenting these technological advances, this talk will discuss possible implications of big data 3D shape and spatial information in archaeology. We will consider the potential for data science approaches to studying the human past, especially applications for machine learning and artificial intelligence.