Abstract: Aspects of the present disclosure relate to systems and methods for adjusting a behavior of an application in real-time. Telemetry for a client computing device, a server computing device, and a network associated with the application may be received. It may be determined, while the application is running in real-time, whether to adjust the behavior of the application using the received telemetry. In response to determining to adjust the behavior of the application, the behavior of the application may be adjusted based on the received telemetry while the application is running in real-time.

Abstract: A user may store content in a data store that includes multiple repositories corresponding to one or more categories, where each of the repositories may include one or more containers for content storage. An inference application may be configured to automatically save the content to a location within the one or more containers of the repository. For example, the inference application may be configured to detect content to be saved for a user. The inference application may analyze one or more attributes of the user and the content to infer a repository category to which the content is related, and determine a confidence level associated with the inferred repository category. In response to a determination that the confidence level is above a threshold, the content may be automatically saved to a determined location within one or more containers of a repository corresponding to the inferred repository category.

Abstract: Relevant files are determined for a user upon detecting a user action such as save, load, open, view, share, or comparable ones associated with a file. The relevancy may be determined through one or more queries based on, a number of criteria, where the queries are executed on local or remote data stores related to the user. For example, files on the local computing device of the user, files in an enterprise network associated with the user, files on a social network subscribed by the user may be evaluated for various relevancy criteria. Files determined to be relevant may be prioritized, ordered, and/or grouped for suggestion to the user and presented through a user interface of an application performing the detected action.