Known-Offender Watchlists
Real-time matching at entrances against your own exclusion list, alerting store security before the aisle.
See Ethical Facial Recognition
Retail and Fuel: Organized Theft Is a Watchlist Problem
Most retail loss is not opportunistic, it is repeat. That makes it a recognition problem rather than a monitoring one, which is why a watchlist on the cameras you already have moves the number.
In Production
Israel, and a Florida Pharmacy Chain
In Israel, IREX facial recognition helps local retail chains prevent millions in losses from organized shoplifting. The system matches individuals against a database of known offenders as they enter stores and alerts security staff in real time. IREX is also deployed in a Florida pharmacy chain.
This deployment shows the platform applying beyond public safety into commercial security, where rapid identification of repeat offenders has a direct effect on the bottom line. It also carries the platform’s constraints unchanged: only pre-registered individuals are recognized, and every privacy-sensitive action is logged against a recorded reason.
The same estate covers the operational side: unattended items, perimeter and after-hours intrusion, fire and smoke, crowd density at entrances, and camera-integrity monitoring so a large multi-site fleet stays working between service visits.
Coverage
Stores, Forecourts, and Back of House
Forecourt and Plate Control
Plate and vehicle-attribute recognition at fuel and EV charging sites, including drive-off and repeat-vehicle detection.
See License Plate RecognitionUnattended Items
A bag or a box left in a public area of the store, raised without alerting on every shopper who sets one down beside them.
See Abandoned Object DetectionAfter-Hours Intrusion
Perimeter and restricted-zone monitoring when a site should be empty, across a dispersed estate, with breaking glass heard on the camera microphone.
See Perimeter Intrusion DetectionFire and Smoke
Early detection in stockrooms, warehouses and forecourts, ahead of point sensing.
See Fire & Smoke DetectionFleet Integrity
A sprayed, knocked or defocused camera raises its own alert at every site, which is how a few hundred stores stay covered between service visits.
See Camera Tamper DetectionOn Camera
The Forecourt at Four in the Morning
Plate, Make and Color at the Pump
A fuel and charging site is a dispersed estate with almost nobody on it, and the vehicle is the identity. Plate plus make, model, color and body type are read at the pump and at the entrance, so a drive-off becomes a record rather than a report and a vehicle that has done it before is recognized on arrival. None of it depends on a new camera: the forecourt cameras already point at the right places.
License Plate Recognition
FAQ
Is watchlist matching in a store lawful?
It depends on your jurisdiction and on the basis for your exclusion list, and that is a question to settle with counsel before deployment. The platform supports it the same way it supports every other privacy-sensitive action: only pre-registered individuals are recognized, the operator records the lawful grounds, and the action is logged in an append-only trail. IREX also runs background checks on prospective clients and declines business it judges high-risk on data collection and security.
Can it run across hundreds of small sites?
Yes, and that is the usual retail shape. Small sites typically use a single cost-effective edge server, larger ones a direct connection, all under one central deployment with one management interface and one audit trail.
Do we need a person watching?
No, and that is the point. Detections alert store or central security with the frame attached, and a person verifies. Nothing responds autonomously. Professional monitoring is available if you want a call center in that loop.
Who owns the data?
You do. Customers retain 100% ownership and control of their data, and IREX neither owns nor accesses customer video, events, logs, watchlists, or floor plans. On-premises deployment is the data-residency mechanism.
How do we start?
With a pilot: one site, three to five use cases, six to twelve weeks, and two or three measurable success criteria agreed in writing before it begins. Accuracy is benchmarked on your own cameras, and the measured numbers go into the contract.
Pilot It in Your Worst Ten Stores
Loss-prevention analytics should be measured where the loss actually is, over a full season if possible.