The direct and indirect costs of cyber attacks continue to mount while the volume of attacks is increasing. Recent attacks against crypto-currency exchanges illustrate the drastic second and third order effects of such attacks. Hence, such incidents are best avoided. Meanwhile, advances in artificial intelligence and threat intelligence hold promise for prediction of specific attacks weeks before occurring. In this talk, we provide an overview of a new technology called DarkMention that leverages these new predictive techniques. This platform, developed under funding by the U.S. intelligence community has been shown to predict specific attacks against an organization with single-digit false positives in blind tests on real-world attack data. This talk reviews the technology, describes various case studies, and discusses how it is being integrated into real-world enterprise environments.