Read the graph
Transactions, stock moves, price changes, promo history, seasonality, calendar — all pulled from the retail data graph with full context per store.
A demand forecast for every SKU in every store, every day — adjusted for drops, seasons, markdowns, and promo lift. One number your whole team works from. Not five disagreeing spreadsheets.
Kwanta Forecasting is opening for early access with multi-store fashion and specialty retailers. Join the waitlist to get first access and help shape the defaults.
Forecasting is the primitive underneath replenishment, promo, assortment, and pricing — one graph, one signal, every workflow.
Transactions, stock moves, price changes, promo history, seasonality, calendar — all pulled from the retail data graph with full context per store.
Category → SKU → store × day. Sensible pooling for sparse SKUs. Cold-start for new launches via analogs and size-curves.
When a promo launches, a stock-out hits, or a price moves — the forecast updates automatically. Every downstream workflow picks up the new number.
Every forecast is explainable — the drivers, the confidence band, the analogs used for cold-start. Your team can override with a written reason.
Category, SKU, store, day. Reconciled top-down and bottom-up so the numbers never contradict each other.
Analogs from similar SKUs, size-curve estimation, life-cycle stage classification — so new launches have a forecast from day one.
Historical lift by promo mechanic and depth, applied forward. The forecast moves when the calendar moves.
Weekly, monthly, annual seasonality. Holiday calendars per market. Drops and collections as first-class events.
Every forecast comes with a confidence interval. Downstream agents read the band, not just the point estimate.
For every forecast: what drove it, what analogs were used, what changed since last run. No black box.
Forecasting sits inside the Kwanta retail graph — no separate data pipeline to maintain.
Talk to the founders. Get early access, honest pricing, and a direct line to the team.