One of the largest US automotive aftermarket parts providers serving professional installers and DIY customers. With nearly 74,000 employees, the client organization operates 5,200+ stores, 100+ branches, and serves stores in the U.S., the U.S. Virgin Islands, Puerto Rico, and Canada.
The client was challenged to transition from 20th century manual retail planning processes to a 21st century digitized process to meet the increased customer expectations of providing an endless aisle of parts with fast availability through disconnected legacy systems and processes.
CoreCompete built a cloud-based analytical solution that senses in real-time, market demands and assorts billions of part/store combinations. Customer needs are fulfilled from inventory across multiple stores while optimizing inventory depth to maximize profit. Analytical solutions drive execution in legacy systems in a highly automated exception-based process.
Forecasting billions of part/store combinations incorporating external demand signals and using 60+ predictors daily
Assortments are optimized daily to provide best market level assortments with dynamic inventory rebalancing through reverse logistics
Automated routine decisions with exception management to improve productivity and enable a shift to weekly/daily cycles from a monthly/quarterly process
Enterprise Data Lake set up in the cloud is a one-stop shop for reporting, data discovery, data science, and analytical apps that drive execution systems
CoreCompete in Action
CoreCompete built an enterprise data lake comprising hundreds of daily feeds of inventory, sales invoices, product lookups, and master data from dozens of disparate and disconnected enterprise systems.
A range of analytical and machine learning models were built to optimize breadth and depth decisions. Demand prediction models account for customer’s willingness to wait, and a large number of key decision variables to create a highly targeted assortment with optimized inventory levels.
The solution was built using SAS Enterprise Miner, SAS Demand Driven Planning and Optimization, Hortonworks, Docker, and AWS services to meet the challenge of elastic, large-scale computing.
CoreCompete delivered the ability to optimize assortments within 3 months of engagement, continuing with analytical application deliverables every three months to complete the entire supply chain analytical transformation within 15 months.
The client felt that the high availability architecture designed by CoreCompete was not only much simpler than the alternatives but also offered them all the capabilities they desired with the lowest AWS costs.