Emerging epidemics are a constant threat to global health. The recent Ebola outbreak in West Africa alone took over ten thousand lives despite international help nearing $5 billion from 70 countries. In retrospect, two factors stand out: (i) the local health system was illprepared, with Ebola-affected countries falling about 80% short of WHO recommendations for numbers of doctors and nurses per capita (223/100,000); (ii) the declaration of the Public Health Emergency of International Concern came late, over 4 months after the first international transmission event. These apparent systemic failures, however, likely reflect the fact that Ebola outbreaks in Africa are so highly unpredictable and variable. Outbreaks prior to the West African Ebola Outbreak occurred in the Democratic Republic of Congo in 2017, 2014 and 2012, as well as Uganda in 2007, but these outbreaks never exceeded 500 cases. In stark contrast, there were nearly 30,000 cases caused by the West African epidemic. The dynamics of an Ebola outbreak are shaped not only by the biology of the virus, but also in large part by societal and behavioral factors. Both factors are heterogeneous and highly variable, as evidenced by contact tracing data from the West African Ebola outbreak. These data suggest that less than 5% of Ebola cases caused more than one secondary infection, yet some individuals transmit the virus to dozens of others. Theoretical models of disease spread are thus becoming increasingly based on stochastic network representations of epidemiological data rather than conventional collections of deterministic well-mixed compartments. We postulate that heterogeneity in behavior is not only an important feature of an emerging epidemic, its presence implies that behavioral interventions can be more important than biomedical interventions in shaping the spread of disease. Here, we propose the development of a modeling framework that combines a stochastic network model with an agent-based model to allow the effects of both behavioral and biomedical interventions on a disease epidemic to be investigated. The network model will provide stochastic forecasting in the form of distributions of possible disease spread outcomes as functions of patterns of behavior. At the same time, agent-based simulations imposed on the network will allow us to quantify the effects of interventions that aim to mitigate disease transmission through alterations in either the behavior of individuals or the biology of the virus itself. These different modeling approaches are thus complementary. Furthermore, each can be independently validated against available data sets. The results of this study will advance our understanding of the modeling and surveillance required in managing infectious diseases, and will have significant implications for public health policy by helping to identify improved strategies for responding to emerging pandemics.