U.S. dams are threatened by age-induced fragility and increased hydrologic stresses due to climate change. In many cases, communities and infrastructure below the dams have also increased dramatically over time, increasing the exposure to dam failure. Given that there are over 90,000 such dams in the United States, a traditional approach to dam risk assessment is challenging to implement. Our I-GUIDE project is taking an integrated approach to the application of "big data" sources so that a national or portfolio risk assessment of these assets can be attempted for the first time. This includes a spatially specific analysis of the climate changes of concern, of what is likely to be impacted if the dams fail, of the cascading effects of those failures on the national economy and other critical infrastructure elements, and the potential resilience of the infrastructure systems given the governance at different levels. The application of machine learning tools, statistical inference, natural language processing and the geo-hypercube together with traditional physics based and economics models is illustrated.