Machine Learning Inventory Optimization Reduces Stock at Uni-Select

Case studies

WMS

Machine Learning Inventory Optimization Reduces Stock at Uni-Select

Case study shows how Manhattan Demand Forecasting and Replenishment with Automatic Policy Tuning cut inventory and exceptions at Uni-Select's Montreal DC.

Published

June 4, 2026

Read time

3 min read

Source

Manhattan Associates

Uni-Select faced seasonal demand swings that drove excess inventory and frequent planner exceptions. Implementation of Automatic Policy Tuning (APT) applied machine learning to continuously adjust forecasting policies across thousands of SKUs. Results included lower inventory, higher fill rates, fewer manual interventions, and improved planner productivity while service levels remained stable.

Key takeaways

Automatic Policy Tuning uses machine learning to refine demand forecast parameters without planner intervention

Seasonal demand fluctuations were addressed by shifting intermittent SKUs to four-weekly forecast updates

Inventory levels declined while fill rates and service levels stayed high

Planner productivity rose as exception volume and manual order review steps dropped

APT is being expanded from the Montreal DC to additional Uni-Select distribution centers

Market overview

SCR methodology note

Vendor landscape

Leaders

Implementation considerations

Important consideration

Original Source

Vendors mentioned in this report

No items found.