Tredence partners with Databricks to co-develop a retail OSA accelerator

Data science and AI engineering company, Tredence Inc., on Monday announced a partnership with Databricks, a company that specialises in unified analytics, to build an On-Shelf Availability Solution (OSA) for retailers.

The companies are co-creating an OSA accelerator that helps retailers and manufacturers solve out-of-stocks by combining Databricks’ data processing capabilities and Tredence’s expertise in AI/ML.

“We’re excited to be working with Tredence as they leverage the power of Databricks’ Lakehouse Platform to help retailers solve critical pain points and uncover actionable insights from their data,” said Rob Saker, Global Industry Lead, Retail & Manufacturing, at Databricks.


Shoppers encounter an out-of-stock (OOS) as often as once in three shopping trips, according to IHL Group research. On-shelf availability (OSA) is the biggest challenge in retail as retailers miss out on nearly $1 trillion in global sales because they do not have on-hand what customers want to buy in their stores, the companies said in a statement.

The OSA accelerator co-created by Tredence and Databricks promotes real-time out-of-stock alerts, enables retailers to optimize on-shelf availability and identifies root-cause drivers of out-of-stock to improve day-to-day sales. The accelerator is a robust foundation of best practices, quick-start guides and notebooks for a full out-of-stock or supply chain solution.

“Together with Databricks, we expect to meet the explosive enterprise customer demand to scale industry-specific AI/ML innovation and deliver transformational industry value in real-time for their business," Rich Williams, VP – Data Engineering, Tredence said.

Enterprise organizations will be able to identify historical out of stocks, predict out of stock risks, create automated and accurate out of stock alert models, equip retail associates to act quickly and replenish shelves while also being able to quantify the business value to improve on-shelf availability.