Cyberattacks are a growing concern in every market with network-connected devices. Attacks such as ransomware, malware, spyware, spoofing, botnets, and more, can result in intellectual property theft, system downtime, reputation damage, and unfortunately loss of life. Most cyberattack detection techniques are limited in scalability and detection efficacy as the cyberattack landscape is constantly widening and evolving. Derivative variants of pre-existing cyberattacks are easily created that bypass existing detection techniques, and the protections have limited visibility into the full cyber system leading to high false-positive rates. Caspia's proposed work, CySense, will develop a real-time hardware-assisted cyberattack detection engine powered by state-of-the-art machine learning algorithms to identify and deter malicious events on network and edge devices. This solution will leverage novel data points and runtime signatures generated by hardware components in modern computer chips to identify when cyberattacks are occurring and proactively prevent them from hurting the victim. This will provide widespread protection against pre-existing and new cyberattacks with real-time results. For Phase I, Caspia will investigate and demonstrate the technical feasibility for novel data processing and machine learning algorithms to identify unique features of various cyber-threats, as well as develop an initial concept model for this technology considering hardware characteristics and emerging approaches in machine learning for anomaly detection. The proposed activities will progress commercialization goals of offering a highly accurate and real-time detection engine for critical cyber threats.