10 AI forecasting models compete per SKU. Newsvendor optimisation calculates profit-maximising stock levels. 13 pre-built data connectors. Single DC or multi-DC.
Newsvendor model factors demand, uncertainty, service level, holding vs stockout cost.
Plain-English per SKU: why this quantity, expected demand, safety buffer.
Change price 5%? Lead time slips? Lower service level? See profit impact instantly.
Tally, SAP B1, Busy, Marg, Zoho, Shopify, Amazon, Flipkart, QuickBooks, PostgreSQL.
Croston and TSB for sporadic, lumpy demand — spare parts, seasonal, new launches.
8 Indian presets: Diwali, Pongal, Navratri, Eid. Historical uplift, forward stocking.
NaiveLast3M — benchmark every AI model must beat.
ETS (Holt-Winters) + SARIMA — trend + seasonality.
Croston + TSB — sporadic, lumpy demand patterns.
RandomForest, XGBoost, LightGBM, CatBoost — with price/promotion features.
Inverse-RMSE weighted. Auto-downweighs poor models.




High SKU count, fast-moving, festival planning.
Intermittent demand, long tail, critical availability.
Expiry-sensitive, regulatory, seasonal.
B2B, variable order sizes, project-based demand.
Single Location: ₹2,499/mo · Multi-Location: ₹4,999/mo. 30-day Pilot — ₹4,999, includes full installation, training, and direct support.