Security monitoring platform for managing access rights associated with cloud applications

A device, comprising:

• one or more memories;
• and
• one or more processors, communicatively coupled to the one or more memories, configured to:
• obtain an access rights data model that is trained based on historical data that includes information related to permissions and attributes associated with multiple users within at least one cloud application;
• obtain, from the at least one cloud application, information that relates to current access rights assigned to at least one user within the at least one cloud application;
• determine a probability that an access level assigned to the at least one user within the at least one cloud application is correct based on the access rights data model and a set of features that relate to the at least one user;
• perform one or more actions to modify the access level assigned to the at least one user based on the probability satisfying a threshold;
• and
• apply a reinforcement learning technique to update the access rights data model based on data that includes feedback on the one or more actions.

A non-transitory computer-readable medium storing instructions, the instructions comprising:

• one or more instructions that, when executed by one or more processors, cause the one or more processors to:
• obtain historical data that includes features related to permissions and attributes associated with multiple users within multiple cloud applications;
• train a data model using multiple machine learning techniques,
• wherein the multiple machine learning techniques include an unsupervised machine learning technique that partitions the features included in the historical data into multiple clusters based on a nearest mean associated with each feature,
• wherein the multiple machine learning techniques include a supervised machine learning technique that derives a mapping function from the features included in the historical data to the multiple clusters, and
• wherein the data model is based on the mapping function;
• determine a probability that a current access level assigned to at least one user within at least one of the multiple cloud applications is correct based on the mapping function and one or more attributes associated with the at least one user;
• perform one or more actions to implement a modification to the current access level assigned to the at least one user based on the probability satisfying a threshold;
• receive feedback indicating whether the modification to the current access level is approved, revised, or rejected;
• and
• apply a reinforcement learning technique to update the mapping function associated with the data model based on the feedback.