A supermarket supply chain exception, played through a team of AI agents in 45 simulated minutes. Step through it, or press play.
One grocery line, one state, 21 to 27 Sep. Daily sales indexed to the trailing six-day average (dashed line = 1.0). REAL
Illustrative figures from the scripted run, not measured results. SIMULATION
| KPI (simulated) | No action | With A + B |
|---|---|---|
| Stores at stockout risk | 6 | 0 |
| Inbound delay felt | 36 h | 22.5 h, no store impact |
| On-shelf availability (zone) | ~91% est. | ~99% est. |
| Extra cost | None | ~$480 |
| Detection to decision | No decision | 20 min |
Real: one product line's daily sales in one state for 21 to 27 Sep, the 1.31× spike, one zone at 1.51×, and a cluster of three stores making up 44% of the day across 12 stores. Simulated: every quantity, order, container, ticket, dock slot, ETA, cost, reply and KPI. The freight forwarder, carrier and 3PL partner are placeholders. A personal experiment, not a system or communication of any retailer.