A modern data platform for retail decisions

A national retail chain’s merchandising teams were making decisions on overnight batch reports — always a day behind the customer, and never fully trusted across 400 stores. We designed a unified, governed lakehouse and built the pipelines and semantic layers that made insight repeatable. Crucially, we worked backward from the decisions teams actually make, killing vanity metrics before they were built and investing in the lineage and quality controls that earn trust. Merchandisers now act on same-day insight across every store, with a platform they trust enough to use before the morning coffee is cold.