A complete theft-detection stack, without the hardware tax.
Vision AI turns the cameras you already own into a theft-detection system. Six detection models, one daily digest, zero new hardware.
Your daily report, every incident — on your phone.
One push notification. A timestamped log of every flagged event from the previous 24 hours, each cross-referenced with the BankID used at entry. No dashboards to babysit, no footage to scrub through.
- Delivered automatically as a daily report
- Tap any event to jump straight to the video clip
- BankID match makes follow-up actionable
Daily Report
Storgatan 12
Tue, June 16 · 4 incidents
Verified events
4
Footage reviewed
24h 00m
Suspected theft
Anna Andersson
SE-7741-XXXX
Unscanned item
Unknown customer
GUEST_TOKEN_18
Loitering
Mikael Berg
SE-2284-XXXX
Vandalism
Erik Svensson
SE-3320-XXXX
From camera feed to daily report.
One continuous loop, running while your store sleeps. No staff in front of a monitor, no real-time alarms to triage.
Existing cameras
RTSP / IP feeds bridged to our EU cluster. No new sensors, no hardware swap.
Vision models
Six detection modalities run overnight: concealment, loitering, group coordination, till tampering, after-hours entry, BankID match.
Daily report
Delivered daily. Every flagged incident, timestamped, video-anchored, identity-matched.
Detection catalogue
Vision models trained on specific retail theft modalities. Tuned to reduce false positives while capturing what humans miss.
Concealment
Items moved into pockets or bags outside of normal shelf interactions.
Abnormal Loitering
Dwell time in high-risk zones or near sensitive equipment like tills.
Group Coordination
Multi-actor scenarios where distraction is used to facilitate stock loss.
Till Tampering
Unauthorized POS access or unusual cash-handling gestures.
After-hours Entry
Any motion detected outside of authorized operational windows.
Entry Log Match
Automatic mapping of visual presence to BankID credentials for accountability.