The pervasiveness of cell phones, mobile applications and social media is generating vast amounts of information that can reveal a wide range of human behavior. From mobility patterns to social connections, these signals expose insights about how humans behave and interact with their environment. While a lot of work has focused on analyzing behaviors, relatively little effort has been dedicated to understanding ways in which such findings could be useful to decision makers in areas like smart cities or public health. In this talk I will discuss two projects: (1) AlertImpact, an agent-based framework that uses geo-referenced cell phone data to model the impact of the preventive actions implemented by the Mexican government during the H1N1 flu outbreak and (2) TweetLand, a method to automatically identify urban land uses and landmarks (point of interest) using tweeting patterns.