Case Study: Lightspeed Restaurant (U-Series) achieves accurate nightly cover forecasts and optimized staffing with Amazon Web Services

Upserve - Customer Case Study

Upserve is a cloud-based restaurant management and POS provider that aggregates payments, reservations, reviews, and menu data to give restaurateurs real-time insights. Facing the challenge of turning historical and streaming data into actionable forecasts for thousands of diverse restaurants, Upserve needed scalable, easy-to-deploy machine learning so each venue could get accurate, customized predictions without heavy engineering overhead.

Upserve used Amazon Machine Learning within its existing AWS stack to build Shift Prep, creating 100+ models that forecast nightly covers and menu popularity (retrained weekly) and deliver predictions in a daily email to owners. The solution went from decision to production in two weeks, improved prediction accuracy over baselines, helped restaurants optimize staffing and menu planning, and is being scaled toward thousands of models and all 7,000+ customers.

Lightspeed Restaurant (U-Series)

Bright Fulton

Director of Infrastructure Engineering