Retail & E-Commerce
Demand Sensing & Inventory Intelligence
AirflowBigQuerydbtLooker
The Challenge
A mid-market retailer with 180 stores and an e-commerce channel was losing $6M annually to stockouts and another $4M to overstock markdowns. Demand planning relied on quarterly forecasts built in Excel, disconnected from point-of-sale and web analytics data.
The Approach
Unified POS transactions, web clickstream, weather data, and local event calendars into a daily demand-sensing pipeline. ML models generated SKU-level demand forecasts per store per day, automatically triggering replenishment orders through the ERP system.
The Impact
Stockouts reduced by 34%
Overstock markdowns cut by 28%
Forecast granularity went from quarterly to daily per SKU
Facing a similar challenge?
Let's talk about how data engineering can drive results for your business.