Prerequisites

In order to provide the most value, Account Defender requires direct access and visibility to traffic for both of our initial account takeover and account creation use cases.

Our machine learning models will learn individual user profiles based upon their typical actions. In order to monitor these accounts and determine fraudulent activities, Account Defender needs context to all relevant actions performed by users in a post-login environment. By providing full visibility to activities conducted post login, we can build an accurate profile of user behavior that will allow HUMAN to determine if specific actions are fraudulent.

Since all environments are different, we offer a multitude of methods by which to define these parameters. Refer to Basic Account Defender Integration to see, which method suits your environment best.

In cases where the fraudulent action happens via a third party (e.g., HyperWallet), we will require some minor additional steps to provide the audit of those actions to HUMAN (e.g. when receiving a call on a web hook, report that to HUMAN via our APIs).

Once HUMAN has access to all relevant traffic in our backend engine, our detection algorithms will go to work.