Product interaction
Surface repeated pickup-and-return behaviour around configured displays.
Retail Intelligence · B2B early access
Your CCTV saw it. Help your team act on it.
Retail teams already have cameras, but most footage becomes useful only after someone searches for an incident. Avirel is being developed to surface selected store events while they may still be actionable.
Pilot-stage conceptBehaviour signals, not shopper identityNo footage storage
Detection scenario
A shirt was picked up, considered and returned. Did anyone learn from it?
In a fashion store, a shopper may pick up an item, compare it, put it back and leave. Standard CCTV records the movement but gives the merchandising or service team no useful signal.
Avirel is being designed to recognise configured interaction patterns around selected racks or displays. The store could use that signal to test placement, notice repeated hesitation or offer timely help.
The camera alone cannot reliably know the exact SKU, whether a sale occurred or why the shopper returned it. That would require integrations such as POS, RFID or inventory data.
Who it is for
For retail teams responsible for the floor, service and safety.
This could support fashion, department, electronics, grocery and speciality retail environments. Brands such as Zara or Lefties illustrate the type of high-footfall store problem—not Avirel customers or partners.
Service opportunity
Alert staff when someone remains in a selected assistance area.
Unusual activity
Flag configured restricted-zone or concealment-like patterns for review.
Queue and safety
Surface selected queue thresholds, possible falls, smoke or unusual movement.
A responsible pilot
One store. One camera zone. One measurable problem.
A broad promise to “understand the whole store” would create noise. A useful pilot begins with a clearly defined behaviour, zone, operating response and success measure.
Choose the question
For example: are customers repeatedly returning a product without receiving assistance?
Test the signal
Review detections with store staff and adjust zones, timing and thresholds.
Measure usefulness
Compare the alerts with actual staff observations and business outcomes.
Useful activity signals without turning shoppers into identities.
Avirel’s current direction does not use facial recognition and does not store store-camera images or footage. Alerts indicate that configured activity may need review; they do not prove intent, theft or wrongdoing.
Activity, not identityFocus on events inside selected zones.
Review, not accusationStaff remain responsible for interpretation and response.
Camera limitsExact product and sale outcomes need additional data.
Bring one store problem worth measuring.
No payment is collected today. We will discuss the camera setup, selected zone and pilot outcome before any early-access test.