AI Inventory Management

The problem

Demand spikes were causing stockouts across 50+ warehouses while other locations sat on dead stock. The buying team was making weekly calls based on a gut read of the spreadsheet.

The system

We built a demand forecasting model trained on 3 years of historical sales, weather, and promotion data, then wired it into the existing Odoo ERP. Reorder suggestions land in a daily digest, the buying team approves or edits, and purchase orders fire automatically to vendors.

The result

Stock accuracy reached 99.5%, holding cost dropped 42%, and the buying team’s weekly review went from 6 hours to 45 minutes.