In this supply chain analytics analysis, we encountered the challenge of working with a dataset consisting of 100 rows and 24 columns. During the Exploratory Data Analysis phase, we conducted quality control checks, assessed supply chain risks, performed Inventory Optimization Analysis using the Economic Order Quantity (EOQ) method, conducted customer segmentation analysis, and explored lead times optimization. However, in the modeling phase, we found that using LightGBM and RNN models was not suitable for this small dataset, resulting in unsatisfactory outcomes. Therefore, we recommend utilizing a larger dataset or considering alternatives such as linear regression, decision trees, or random forests to achieve better results in cost optimization and demand forecasting.In this supply chain analytics analysis, we encountered the challenge of working with a dataset consisting of 100 rows and 24 columns. During the Exploratory Data Analysis phase, we conducted quality control checks, assessed supply chain risks, performed Inventory Optimization Analysis using the Economic Order Quantity (EOQ) method, conducted customer segmentation analysis, and explored lead times optimization. However, in the modeling phase, we found that using LightGBM and RNN models was not suitable for this small dataset, resulting in unsatisfactory outcomes. Therefore, we recommend utilizing a larger dataset or considering alternatives such as linear regression, decision trees, or random forests to achieve better results in cost optimization and demand forecasting.