Cloud-based mapping technologies are changing the way that the world interacts with GIS. Technologies that allow for aggregate querying of data that is both geospatial and temporal presents unique challenges and fruitful lines of inquiry. At CartoDB, we are pushing ahead with new ways of looking at spatio-temporal data visualization--which we have named Torque--, with intriguing results for both scientists and journalists. In this session, we will present use cases that offer unique ways of looking at data. We will also present challenges that lie ahead with our unique technology. My background in mathematical physics studying timeseries analysis has led to interesting insights and crossovers with the developers/hackers that originated the underlying technology. I hope to present the many lessons we've learned from Torque.