For a large part manufacture company in the transportation industry, maintaining the optimal inventory level in their warehouses is crucial to its bottom line. When too many parts are produced and stored, it costs the company excessive financial investment and previous warehouse spaces. On the other hand, if not enough parts in the warehouses, customers will become dissatisfied when orders may not get fulfilled in time. Thus, there are two conflicting goals to balance when planning the inventory: reducing inventory value and increasing customer satisfaction. The optimal strategy is to find the sweet spot of inventory level for each individual part that is most economical and maintaining high level of customer satisfaction at the same time. In a recent project that Dr. Jay Zhou has preformed, he is able to reduce the inventory level for his client company by $16 million and still maintain the same level of customer satisfaction. This work is highly received by the client. In this project, Dr.Zhou takes advantage of machine learning models and reduces huge number of parts to a much smaller number of homogeneous groups. The "Demand Satisfaction" are calculated for these groups.