October 31, 2017 -
Zighra has received its latest patent with the U.S. patent office, enabling the company to develop highly personalized user-device models to safeguard against account hijacking, social engineering, and automated bot attacks negatively impacting the connected economy.
The patent expands on the behavioral intelligence layer of Zighra’s foundational patent on N-dimensional authentication.
It covers Zighra’s newly-created proprietary lightweight machine learning algorithms that can operate locally on a device.
By using device sensors including touchscreens, accelerometers and gyroscopes, Zighra’s machine learning technology invisibly authenticates both smart devices and users.
The technology provides continuous proof of presence based on how the user interacts with their devices including swipe, scroll, tap, walk with — enabling a frictionless user experience.
“Traditional machine learning approaches take a long time to learn and need to run in the cloud or on servers. It is critical that security solutions learn quickly and apply that learning for authentication and threat detection,” said Deepak Dutt, CEO of Zighra. “This essentially requires that we have the capability to implement machine learning and anomaly detection completely on-device.”
The company’s previous patent covers technology that gives it the ability to create cognitive signatures to authenticate each user through sensor-based intelligence obtained from six distinct layers of data.
Earlier this month, ighra launched SensifyID, a patented solution that delivers a 360-degree view of user interactions and continuous proof of identity while preserving the privacy and security of the end consumer.