Ward (GetWard)

Ward (GetWard) — Self-checkout terminal event and loss-prevention dataset: per-transaction intervention and anomaly events at SCO lanes (skip-scans, produce and bakery PLU weigh mismatches where a premium item is keyed as a cheaper code), weighed-item versus keyed-PLU pairs, attendant intervention rates, and per-store net-efficiency metrics reconciling labor hours displaced against incremental shrink applied to SCO sales; adjacent store-level grocery price and promotion optimization data. Micro-scale vendor (2 employees, no disclosed funding) but positioned on a structurally valuable dataset:…

Ward (GetWard) offers (Alternative, Sentiment, Supply Chain) — Bounded by pilot scale - estimate order 10-500 stores under monitoring, which at typical self-checkout volumes means order 1-50 million SCO transactions monitored per year with an intervention or exception rate of roughly 1-5 percent, so order 100K-2M labeled anomaly events per year.

Training computer-vision and sensor-fusion models to detect self-checkout theft versus honest error (a long-standing open labeling problem, since the source explicitly notes the honest and dishonest cases are indistinguishable from receipt data alone); produce PLU misidentification models from weighed-versus-keyed pairs; retail loss-prevention risk scoring per store and per basket; labor-versus-shrink economic modeling for front-end staffing; grocery price elasticity and markdown optimization

Coverage spans US; Application Software, Food Retail; alternative, sentiment, supply_chain; equities.