Examples are described herein for facilitating responses to computing device service requests. In various examples, natural language processing may be performed on a plurality of incoming computing device service requests associated with a plurality of computing devices. A plurality of embeddings corresponding to the plurality of computing device service requests may be generated based on the natural language processing. Based on distances between each of a particular subset of the embeddings in an embedding space, a trending service issue associated with the computing device service requests corresponding to the particular subset may be predicted. The trending service issue may be provided to a computing device servicer associated with resolving a new incoming computing device service request.