Custom top-down probabilistic finished goods forecasting reconciled with AI driven bottom-up material estimates
An industrial products manufacturer migrated from selling a standard set of SKUs to allowing mass customization of the product portfolio by the customers. The requirement made it very difficult to forecast requirements at the finished goods level. A probability based configurator input at the options level was leveraged to estimate WIP and raw material inventory at a product family level. A corresponding custom bottom-up forecasting at a component level across families was leveraged to reconcile the inventory requirement using machine learning predictive techniques. Rapid customization, implementation and integration with the Client’s ERP system ensured seamless absorption and effectiveness from Day 1.
- Real time demand sensing and responding to material consumption patterns
- Decreased material stockouts by more than 30%
- Compliance to customer quoted leadtimes increased from the low 70s to the low 90s
- Substantial reduction/elimination of manual imports and exports of data across legacy planning systems, ERP and CRMs