Data

Every figure in Vantage — SKUs, regions, inventory, twelve months of demand history, supplier lead times — is synthetically generated from a fixed random seed when the page loads. Nothing here is a real company’s data; the generator is written to be internally consistent (seasonality by category, lead time by sourcing region, cost by product type) so the mechanics of the tool demo realistically.

Forecast method

Each SKU × region series is de-seasonalized using a category-level seasonal index, fit with a short linear trend over the trailing six months, then re-seasonalized forward three months. It is intentionally simple — built to demonstrate the workflow, not to be production-grade demand planning.

Risk scoring

Coverage days = (on-hand + adjusted in-transit inventory) ÷ forecast daily demand. A SKU × region is flagged Critical when coverage falls under 75% of the effective replenishment lead time, High under 110%, Medium under 175%, and Overstock above 120 days of cover.

Scenario modeling

Each disruption carries a base effect (added lead time, a supply reduction, and/or a demand shift) applied only to the SKUs it plausibly touches — by sourcing region or product category — and scaled by the severity slider (0–150%). Effects from multiple active disruptions stack, capped to stay within a plausible range.

Deployment

This build runs entirely client-side — the generated dataset is the single source of truth for every chart, table, and KPI on the page. A static export of this file is the deployable artifact for hosting on Hostinger, GoDaddy, or any static host.