The goal of this Capstone project was to use Azure machine learning and choose a set of models that would accurately predict the resource usage and response time values for a set of UPS transaction types. These transaction types are for an application that has very high workload volumes and executes on a large number of servers.