Albertsons is working to enhance merchandising intelligence at scale for its grocery retail businesses in the field.
The grocery giant is working directly with merchants, or store operators, to transform multiple platforms into a single space for decision-making with improved product, pricing, promotions and placement insights, said Karthik Iyer, group VP, head of merchandising, transformation and AI, at Albertsons.
The merchandising intelligence platform is built on a Databricks foundation, Iyer told CIO Dive. The grocer is in the process of scaling the platform to merchants with the goal of having the product fully rolled out by the end of 2026.
“We are shifting the merchant behavior of how they’ve been used to working in tech into a completely different thinking space,” said Iyer, who reports to Anuj Dhanda, Albertsons’ chief technology and transformation officer.
Merchandising intelligence is one of four key investments Albertsons has planned for this year as it looks to scale AI across the enterprise and compete in an intensifying race to digital leadership among grocery stores.
Building an AI-first platform
Iyer said it was crucial to build a platform that incorporated clean data, governance and AI.
The Databricks’ Lakehouse offering houses Albertsons retail data, Iyer said. The next level of the platform is the governance layer through Databricks’ Unity Catalog and AI Gateway. Sitting on top of the stack is Databricks’ AI agent Genie, which enables merchants to easily query the platform in natural language.
Incorporating large language models into the platform to train on clean retail transaction data in Lakehouse, such as price points for green apples versus red apples over the last three years, allows the platform to provide a clear, forward-looking diagnosis, Iyer said.
Genie lets merchants tap into the value of the data within the platform, Iyer said.
“If we have a dry summer and it’s not as sunny as it was the previous year, what will it mean for Tillamook ice cream sales, do I shrink my space, do I grow my space, do I add more flavors, or do I reduce the space and give it to something else?” Iyer said. “That’s the kind of insights we are able to draw from this combined platform with Databricks.”
Merchant adoption of the platform will be key to success, which is why Iyer said he worked with the retail businesses to build the system alongside them rather than purchasing an off-the-shelf tool.
“If our merchants are able to say ‘Hey, it’s giving us insights we haven’t thought about, it’s giving us deal economics we didn’t have the headspace to sit down and think about,’ that is what I get excited about the most,” Iyer said, “transforming our business for a meaningful outcome.”
Albertsons saw a need for better real-time understanding of the business, which is where Genie in particular played a strong role in the grocer’s merchandising intelligence platform, said Ken Wong, senior director of product management at Databricks.
“They wanted to provide a real-time, any-type-of-question system that could meet these needs,” Wong said. “With the advances in LLMs, it was possible to build these systems, but having it produce results that were reliable and trustworthy was actually quite an endeavor.”
