Whereas schools typically record mounds of data regarding student performance, attendance, and other behaviors over the course of a school year, rarely is that data consulted and used to inform day-to-day instructional practice in the classroom. As teachers come under increasing pressure to ensure success for all of their students, we are attempting to provide tools to help teachers make sense of what is happening in their classrooms and take appropriate proactive and/or remedial action. One such tool is a Web service we've dubbed the Classroom Sentinel. The Classroom Sentinel mines electronic gradebook and other student information system data sources to detect critical teaching and learning patterns and bring those patterns to the attention of the teacher in the form of timely alerts. In this paper, we introduce the notion of classroom patterns, present some examples, and describe a framework for alert generation and delivery.