Use cases¶
The demo supports several retail stories, but their current maturity differs.
| Use case | Current evidence | Status |
|---|---|---|
| Real-time POS and product performance | Receipt, line, payment events; KQL aggregates; Direct Lake report | Implemented core story |
| Inventory signals and replenishment | Inventory, stockout, and reorder events/tables | Implemented with normalized, contract-tested current-state semantics |
| Customer presence and in-store movement | Entry, Bluetooth Low Energy (BLE), and zone-change events | Implemented synthetic scenario |
| Omnichannel fulfillment | Online-order headers, lines, picked, and shipped events | Implemented with documented live/history differences |
| Supply-chain dwell | Paired truck lifecycle, Eventhouse SLA function, Silver/Gold dwell, queryset and rule | Cross-layer contract tested; live Fabric validation pending |
| Marketing attribution and promotion ROI | Seven-day last-touch journeys, reconciled promotions/payments, campaign return on ad spend (ROAS), and purchase-level audit | Implemented and contract-tested |
| Machine-learning insights | Required forecast/churn/segmentation/stockout plus optional delivery and experimental pricing outputs | Required Reporting outputs are contract-gated and live-validated in standard/full-demo |
| Ontology and conversational agents | Business-entity ontology and semantic/ontology agents | Optional, capability-gated, and live-validated for full-demo |
Core presentation journeys¶
Store operations¶
Start a bounded live stream, show minute sales and store activity in Eventhouse, then show durable Silver/Gold data and the store-operations report page.
Inventory and replenishment¶
Show inventory movements, reorder signals, stockout detections, and current inventory Gold output. Label detections and snapshots accurately; do not call a recent detection "open" without state folding.
Omnichannel¶
Follow an order from creation through picked and shipped events, then compare store and online sales in the semantic model.
Customer journey¶
Use synthetic entry, BLE, and zone-change data to discuss store flow and dwell. State clearly that identifiers are generated synthetic demo data, not real customer records.
Optional enhancement journeys¶
- Governed pricing approval and writeback: ENH-002
- Persona-specific agents and Copilot: ENH-003 and ENH-004
- Supplier collaboration, embedded analytics, and retail media are product ideas, not current deployed capabilities.
- Deterministic weather, holiday, vendor, and customer-state scenarios: ENH-009
The traceability ledger is the authority for implemented and verified status.