Prioritizing messages for server processing based on monitoring and predicting server resource utilization

Aspects of the disclosure relate to prioritizing messages for server processing based on monitoring and predicting server resource utilization. A computing platform may monitor a plurality of servers (e.g., processing engines) to determine a current processing capacity and a maximum processing capacity of each processing engine. The computing platform may predict a quantity of application messages that may be received and/or processed by each processing engine at a later time. The computing platform may generate an application message prioritization framework comprising the prediction, the current processing capacity, and the maximum processing capacity of each processing engine. The computing platform may use the application message prioritization framework to prioritize received application messages to be processed, and to determine whether an application message should be stored in a backlog for processing at a later time or whether the application message should be transmitted to a processing engine for processing.