Abstract: A computing system and method for training one or more machine-learning models to perform anomaly detection. A training dataset is accessed. An overall sensitivity score is determined that indicates an amount of sensitive data in the training dataset. Machine-learning models are trained based on the training dataset and the overall sensitivity score. The machine-learning models use the overall sensitivity score to determine a threshold. The threshold is relatively low for datasets having a large amount of sensitive data and is relatively high for dataset having a small among of sensitive data. When executed, the machine-learning models determine if a probability score of features extracted from a received dataset are above the determined threshold when a second overall sensitivity score of the received dataset is substantially similar to the overall sensitivity score. When the probability score is above the determined threshold, the machine-learning models cause an alert to be generated.

Abstract: Some embodiments manage storage of access data to provide flexible and granular control over storage costs without risking policy compliance, regulatory compliance, or data breach investigation. Resources are classified and given metadata labels. Resource access data is associated with the accessed resource metadata label. A mapping is defined between metadata groups and access data storage boxes. Access data storage box definitions may specify metadata labels. A mapping structure also defines a policy governing use of available storage capacity in access data storage boxes. Per the policy and the available capacity, particular access data may be stored in a particular box, be spilled over to a different box, or be denied storage. Accordingly, the costs of storing access data can be capped and made predictable, and storage of specific kinds of access data can be favored.

Abstract: A method includes receiving audio at a computing apparatus (104), determining, by the computing apparatus, whether the audio corresponds to a predetermined mapping between an utterance and a software call of software application executing on the computing apparatus, and invoking the software call only in response to the audio corresponding to the software call, wherein the invoked software call at least one of activates or deactivates at least one of a mode or a tool of the executing software application.