Most compromised identities show a deviation from the user’s normal behavior in their interactions with the Identity Provider. It may seem easy to detect compromised account by using simple rules based on past behavior, but these simple rules generate too many false positives, and the signal gets lost in the noise. Come get a peek under the hood of the data science and how behavioral analytics, machine learning and other risk signals can be used to design high-quality identity compromise detection algorithm working in real-time. (AI, Machine Learning, Security, Dana Kaufman)