SBIR-STTR Award

Revelio is a User and Entity Behavior Analysis (UEBA) system that uses machine learning techniques to detect anomalous user behavior. Revelio works off logs already collected by the Army in an Elastic Stack system. Revelio employs multiple anomaly detectors working off different types of logs to increase the likelihood of detecting anomalies. It uses reinforcement learning techniques to reduce false-positive rates. Revelio is also able to raise the level of ion seen by operators from low-level detections to higher-level attacks.