Abstract: Context information, indicative of a tenant's engagement with a multi-tenant service, is obtained. An engagement state for the tenant is determined and the context information, and engagement state, are correlated to an issue to be addressed. A user experience is conducted, based upon the likely issue to be addressed.

Abstract: Audience targeted filtering of content sections is disclosed. A content filter application initiates operations to provide content filtering based on a detected audience upon receiving a request to provide a content. The content includes audio/video streams and/or text based content, among others. Next, an audience category to provide the content is identified. If the content is available for priori analysis, the content is analyzed to detect section(s) of the content to filter based on the determined audience category and associated filtering rules, and results of the analysis are provided. If the content is not available for priori analysis, the associated filtering rules are applied to detect the section(s) of the content to filter. Furthermore, the filtered content is provided to the display device.

Abstract: Context information, indicative of a tenant's engagement with a multi-tenant service, is obtained. An engagement state for the tenant is determined and the context information, and engagement state, are correlated to an issue to be addressed. A user experience is conducted, based upon the likely issue to be addressed.

Abstract: An issue is identified based on context information and a state of engagement of a user. The issue is matched against an agent based on agent capabilities exposed by the agent. Feedback information is used to generate a reputation metric for the agent.