Real-Time Inventory Management Using Hadoop

Inventory management not only tracks the purchasing, selling and availability of items but also involves the predicative strategies to create a purchasing plan that will ensure that items are available when they are needed in an appropriate quantity (that neither too much nor too little in storage). Further, real-time inventory management gives more strength to the system, which means changes to the inventory including the addition or removal of an item are reported to the item tracking system in real time, i.e., single point truth is to be maintained. In today’s digital era, online marketing is growing with a rapid pace and dozens of terabytes of data is generating per day, which leads to incorporation of big data analytics in inventory management. So many smart companies are now moving toward big data analytics to improve warehouse inventory management. Hadoop is a distributed framework for structuring big data and make it useful for analytics purposes. The Hadoop ecosystem gives the ability to create value from its data by being able to process and store vast amount of data from disparate sources and enables faster processing on larger datasets. In this paper, we discussed about the tools which can be used to migrate point of sales data from different legacy systems to Hadoop and how Hadoop will process these data to maintain single version of truth. The architecture and algorithms used by the Hadoop to provide real-time analytics have also been discussed.