Retail

+18% product availability with shelf-image analysis

Computer vision that analyzes shelf photos from stores to catch out-of-stocks and misplaced items.

The problem

Manual shelf audits caught stockouts hours or days too late, costing lost sales across the store network.

The solution

We deployed a fine-tuned detection model over store photos, streaming stockout and misplacement alerts to store-ops dashboards.

+18%

Product availability

150+

Stores monitored

−70%

Audit time

Stack

YOLOv8ONNX RuntimeEdge TPUMQTT