Résumé : Techniques for making future predictions based upon the present and past data, has always been an area with direct application to various real life problems. We are discussing a similar problem in this paper. The problem statement is provided by Kaggle, which also serves as an ongoing competition on the Kaggle platform. In this project, we worked with a challenging time-series dataset consisting of daily sales data, kindly provided by one of the largest Russian software firms - 1C Company. The objective is to predict the total sales for every product and store in the next month given the past data. In order to perform forecasting for next month, we have deployed eXtreme Gradient Boosting (XGBoost) and Long Short Term Memory (LSTM) based network architecture to perform learning task. Root mean squared error (RMSE) between the actual and predicted target values is used to evaluate the performance, and make comparisons between the deployed algorithms. It has been found that XGBoost fared better than LSTM over this dataset which can be attributed to its relatively higher sparsity.