Methods, systems, and processor-readable media for remotely providing a device status alert. In an example embodiment, data indicative of the status of one or more devices can be subject to an HMM (Hidden Markov Model) and a dynamic programming algorithm to determine the latent state of the device (or devices). A status alert model can be trained based on such data and can be expanded with respect to a wide range of devices including utilizing semi-supervised learning. The alert status model can then be integrated into a device management application that provides a status alert regarding one or more of such devices based on the status alert model.