Inventory accuracy and sales prediction​


Using Azure Databricks, Blueprint’s Data Scientists predicted inventory counts on key SKUs across thousands of stores and uncovered patterns of inventory mismanagement by store associates to save time, money, and resources.

Our work

  • Data science
  • Azure Databricks
  • Machine learning
  • Advanced analytics
  • Power BI

The Problem

A nationwide discount retail company struggled to maintain inventory accuracy and manage “shrinkage” across its retail locations. With inventory checks conducted only once a year, a lack of shipment and unique SKU tracking from the distribution center to a given store, and shrinkage due to product damage and theft, they were unable to accurately manage inventory at many individual locations. Having recognized the need to adopt advanced inventory-control, stock-replenishment, and stock-adjustments/shrinkage controls using advanced analytics, they turned to Blueprint for help.

The Blueprint way

Blueprint leveraged Azure Databricks and Power BI to deploy advanced analytics and machine learning with the client’s sales and operations data. In just a few weeks, Blueprint created machine learning models that performed on par with historical models at a predicting perpetual inventory in addition to a very accurate sales-forecasting model. The output enables the client to see a virtual inventory count at any time for any given store, reduce unnecessary spend, and better manage inventory and shrinkage.


  • Developed and trained regression models to predict inventory counts at over 85% accuracy levels on the model’s first pass.
  • Drastically improved sales forecasting at the SKU/store level with over 90% accuracy using store and customer demographic data.
  • Identified hidden patterns of poor stock-replenishment practices at specific stores, alerting management to training opportunities.

Let's build your future.

Share with your network

Share on twitter
Share on facebook
Share on linkedin
Share on email

You may also enjoy

Case Study

Legacy data migration with a custom data factory

Building a data factory to enable analysis of terabytes of legacy data

Case Study

Supply chain monitoring and predictive analytics

Connecting disparate systems for end-to-end supply chain management