The purchasing department of Thenamaris has formulated a three-prong framework to encompass Machine Learning into boosting operational efficiency and effectiveness. Via unsupervised, supervised and prescriptive learning we are able to get a high degree of visibility into the key drivers behind business decisions, predict future outcomes and optimize end-to-end cost functions respectively. Indicatively, we have used clustering to identify critical provisions, consumables and spare parts, statistical forecasting to project the tactical planning needs of these critical items and optimization algorithms to reach a consensus forecast that yields the perfect balance between vessel replenishment needs versus forwarding, inventory and stock out cost minimization. The cost savings are accompanied by a reduction in operator admin workload through the furthering of digital procure-to-pay process automation.