Sales forecasting is critical to business success, laying the foundation for anticipating future demand and improving inventory management and marketing strategies. Recognizing the importance of sales forecasting, this project leverages a dataset from Kaggle's “Walmart - Recruitment Prediction Competition” covering weekly sales from February 5, 2010, to November 1, 2012, and uses XGBoost, a powerful machine learning algorithm known for its efficiency and accuracy, combined with optimization techniques to achieve optimal performance. We aim to deliver a scalable and accurate sales forecasting solution that supports data-driven decision-making in the retail industry and beyond. The accuracy achieved in our project significantly outperformed the top-performing models from the Kaggle competition using the same dataset.