Estimating energy consumption in cement mills is critical for the cement industry. Following data science practices and adopting machine learning technologies, SYMBIOLABS designed and implemented energy consumption prediction models for a cement mill of TITAN Inc. plant in Kamari Viotia. The models exploit historical milling process sensor measurements and operational data (e.g., raw material ratio, motor loads, output qualitative and quantitative measures, etc) and give prediction for energy consumption with accuracy better than almost one order of magnitude compared to models currently used by cement industry based on bibliographic methods.