Case studiesRetail & venues

People-flow analytics without storing a single face

Venue and retail operator

Headline result
0images

stored, on cameras already installed

0images

Identifiable images stored, by design

34cameras

Existing cameras reused, no new hardware

<1s

From camera to live occupancy dashboard

IndustryRetail & venues
RegionEU
ServiceCustom AI Systems
DeploymentEdge, on-prem
TimelinePoC 4 weeks, rollout 10 weeks
StackEdge inference, computer vision, streaming, dashboards

The challenge

The operator needed live occupancy and flow across its spaces, but could not send video to the cloud or keep identifiable images under GDPR.

What we built

We built a computer-vision pipeline that runs on edge devices next to existing cameras. Video is processed on the device and only anonymous counts and flows leave it.

The result

Live occupancy across every space, with no personal data stored and no new cameras installed.

“Privacy was the requirement that killed every other proposal. Here it was the starting point.”