The time Series Forecasting initiative addresses a critical operational challenge faced by a client in the insurance sector—namely, the escalating processor consumption within a Mainframe computing environment. Anchored in the Cross Industry Standard Process for Data Mining (CRISP-DM) framework, the project adheres to a rigorous methodological progression encompassing business problem articulation, data exploration, preprocessing, and advanced modeling. A diverse array of time series forecasting algorithms is systematically evaluated through empirical experimentation, with each model assessed against stringent business performance metrics. The outcome is a suite of robust, deployable forecasting solutions that not only elucidate temporal patterns in system usage but also enable proactive resource management within the client’s enterprise architecture.