Perishable Replenishment

Simple demand forecast, committed-PO guardrails, zero engineered spoilage

Select a Demand Scenario

Pick a scenario, then run the replenishment plan. The forecast is a trailing 14-day weekday-adjusted mean and nothing more; the value lives in the guardrails that turn that blunt forecast into a buyable, non-spoiling purchase order.

Forecast Samples

Representative store and SKU histories over 90 days. The shaded band is the trailing 14-day window the forecast averages; the highlighted dot is the resulting forecast point (weekday-adjusted). Promo days are marked. This is a plain trailing mean, not a learned demand model.

Run a plan on the Scenario tab to see forecast samples.

Order Proposal

Every store and SKU line, raw forecast cases through the guardrail chain to final cases. Click a line to see the waterfall of guardrail steps. Filter by store or temperature class.

StoreSKUTempShelf (d) Forecast/dayRaw cases Final casesStatus

Guardrails

Six committed-PO guardrails run in order on every line. Counts show how many lines each rule actually changed in the current run.

Run a plan on the Scenario tab to see per-rule counts.

Waste vs Stockout

The honest trade. A naive order-up-to policy hits zero stockouts, but only by overstocking perishables into thousands of cases of spoilage. The guardrails eliminate that waste and accept a modest, deliberate stockout risk on the shortest-shelf lines.

Run a plan on the Scenario tab to compare.