Due to their convenience and effectiveness, unmanned retail stores have grown in popularity in recent years. However, the success of these markets depends on accurate sales and demand forecasting. In this paper, a machine learning-based method for forecasting demand and sales at unmanned retail stores is proposed. More specifically, we forecast demand and sales for various products in the unmanned retail market using a combination of time series analysis and machine learning algorithms. The success of these markets, which run without any staff and depend on technology to make the shopping experience easier, is highlighted by this project, which emphasizes the significance of precise sales and demand forecasting. The system will use historical sales data and real-time data to predict future sales and demand for different products which can be used in optimization of inventory management. This results in minimizing the stockout or overstocking risks and beforehand prediction of the trends. The proposed method, which uses machine learning algorithms to forecast sales and demand for various products in the unmanned retail market, is briefly described. It is also mentioned that the evaluation of the strategy using actual sales data demonstrated high accuracy in predicting sales and demand.