Databricks Genie retail agent
A governed Databricks Genie agent that lets retail teams explore enterprise lakehouse data in natural language.
Problem
Retail teams waited on analysts for every data question. The bottleneck slowed assortment and space decisions across a global store network.
Approach
As the team lead, I conceived and sponsored a Databricks Genie agent on top of the enterprise lakehouse, and led the team that built it. It is grounded on curated gold-layer tables in Unity Catalog (a medallion bronze/silver/gold architecture), so natural-language questions become governed, trustworthy queries over certified business data. It goes beyond simple answers to advanced end-to-end analysis and visualizations: conversational analytics and business intelligence in one place.
Impact
Adopted by the Retail Merchandising, Category Management, and Retail Space Management teams for day-to-day decisions, reducing analyst bottlenecks and shifting business users from reactive reporting to self-serve insight on governed lakehouse data.
Built on Databricks Genie.