Machine Learning powered Adaptive Authentication

Have you ever heard about Machine Learning powered Adaptive Authentication or how we call it – “Multifactor Authentication for the lazy”?

Joke aside, nothing wrong with being lazy in this regard. For us at ThingsRock it just means providing a smooth user experience combined with high security standards.

Imagine a user only needs a second authentication factor (like a PIN or a hardware key) when logging in from an unknown location or from a new device based on their previous login activity.

Our Risk Assessment Engine does exactly this all automatically.

It maintains individually trained (by Machine Learning) login behavior models for the user to detect anomalous or unusual logins.

When a user authenticates against our Identity and Authorization Service, our Risk Assessment Engine evaluates their individual login behavior to determine whether it is a typical login for the specific user or not. This is accomplished by deriving more than sixty individual data points from the current user’s login procedure and use those to query our Risk Assessment Engine in real time through its API.

If our Risk Assessment Engine detects an unusual behavior, it will typically ask for a second factor.

But in case you prefer to just monitor logins or outright block the login, then this is possible, too, as our Risk Assessment Engine fully integrates with our configurable login flows.

And as the login behavior models are regularly re-trained by our machine learning pipeline our Risk Assessment Engine will adopt to new user login behavior automatically.

All of this happens in the background, maintaining a smooth login user experience combined with the high security provided by Multi Factor Authentication. If you would like to know more or are interested in using our Identity or Authorization service, then feel free to contact us.