From multichannel demand to a purchasing plan people can use

The operating problem

A sales forecast alone could not answer when to buy, which channel needed the inventory, or whether an incoming shipment would arrive in time. Spreadsheet models and operational screens had to agree on product identity and on the supply already committed. A proposed purchase could not be treated as inventory that already existed.

What we built

We built a Python forecasting and material requirements planning engine that translates channel demand into dated inventory projections and purchasing recommendations. It brings together sales, available inventory, confirmed incoming supply, product attributes, and planning assumptions maintained in Google Sheets.

The engine supports local Excel review and a hosted workflow backed by MySQL. Operators can inspect a run, compare it with a no-new-orders scenario, review shortages and recommendations, and choose the plan used by the shared operating screens. Shopify and Amazon retain their own demand assumptions rather than borrowing one channel’s sales pattern for the other.

What the work enables

The delivered platform connects the forecast to an inspectable buying decision: what creates the shortage, which receipt changes the position, and which recommendation addresses the gap. Scenario outputs and published decisions have distinct roles, so a planning experiment can be reviewed before it becomes the operating plan.