Tim Greene of Network World on new startup Cyber adAPT using predictive threat analysis to thwart cyber attacks:

Cyber adAPT, a startup springing from DARPA funded research, is shipping its first products that detect network compromises and gather data that can be used later for forensic analysis of breaches.

The company’s appliance-based platform monitors network traffic looking for suspicious communications that might indicate a breach and correlates it with threat feeds to improve its accuracy.

The company just landed a $4.1 million Series A round from Alvin Fund, Granite Point Capital Partners, Griffin Fund II, and Fundamental Capital Management. It started work about 15 months ago and grew out of Irvine Sensors, which had done research funded by the Defense Advanced Research Projects Agency.

The goal is for the platform to perform predictive threat analysis in which it would determine where threats are moving, the goals of the attackers, where the attack is likely to pivot and what assets it is likely to pivot to, and what phase the attack is in.

Predictive threat analysis has long been thought of as the holy grail of cyber security since it would allow networks to dynamically modify policy to account for potential new threats. However, throughout the years it has been very difficult to nail down, often-times producing highly spotty results leading to no better security posture than without the technology.

Maybe this startup has a unique and novel method that has cracked the predictive analysis code?