Fuel demand prediction for an oil company with machine learning methods with self-learning capabilities. BCS was responsible for the design of relevant data interfaces and data control layer, delivery of predictive method based on competitive comparison of several machine-learning models, the dashboard based reporting of the prediction outcomes with Power BI.

At least 50-70% more precise prediction on average than the current model available, and the solution applies self-learning techniques which enable increase of efficiency by analysing the numerous factors by using self-learning which leads to further accuracy of the prediction model.