In an interconnected and fast-moving world, effective policymaking increasingly relies on understanding complex human networks and behaviors. For example, pandemic response requires understanding how disease spreads through contact between individuals and how individuals alter their behavior in response to policies and disease. My research develops computational methods to infer robust signals about networks and behaviors from data, then leverages these signals to tackle societal challenges from pandemics to polarization to supply chains. In this talk, I'll discuss one line of research on mobility networks and disease modeling, including (1) inferring fine-grained mobility networks from aggregated location data, (2) modeling the spread of COVID-19 over mobility networks to inform reopening strategies and analyze disparities. This talk will focus on the two works below, and also touch on our related works in KDD'21 and AAAI'23. Papers: 1. Chang*, Pierson*, Koh* et al. Mobility network models of COVID-19 explain inequities and inform reopening. Nature 2021. 2. Chang*, Qu*, Leskovec, and Ugander. Inferring networks from marginals using iterative proportional fitting. Under review.