The objective was to predict the monthly demand of products needed in the distribution centers of a well known argentine company (real data). To make the prediction we had the sales records from the previous two years (+5.5M entries) and during the course of the year, with the help of our mentor, we conducted all the stages of a Data Science project: understanding the problem; cleaning, exploring and analysing the data; feature engineering, building predictive models, evaluating them; visualizing and presenting the results