The objective of this project is to minimize wastage of meal kits in retail stores. Currently, this is being done by tracking each individual item from the source until the point of sale. This is a cumbersome process and is labor intensive. In order to realize the objective using machine learning the first step in the process is to have an accurate forecast of the demand. This project focuses on generating accurate forecast for each individual item (46 unique items) for each store (47 unique stores).