A camera estate without analytics is an archive. It tells you what happened after somebody already knows something happened, and it depends on a human watching the right one of forty screens at the right moment. Nobody does that reliably for eight hours, and no control room is staffed to try.
The problem
Recording is not watching.
Software over your existing estate
Dragonfruit AI analyses the feeds from IP cameras you already have. No chain-wide hardware replacement, no vendor lock-in to one manufacturer’s model, no capex per camera.
Objects, behaviour, and the exception
Sub-second detection on live streams: intrusion, loitering and dwell time, shelf sweeps, vehicle and plate recognition, and watchlist matching — raised as events, not recordings.
Item recognition at the point of sale
AI can identify the physical item at the scanner and reconcile it against what was rung up, flagging mis-scans, substitutions and price overrides as they happen.
The same feed answers other questions
At Kleinfontein the analytics layer is integrated with the weighbridge and scales, so the platform that detects intrusion also reports on production movement.
What a verified event looks like.
Where it fits
Where it fits.
Analytics are the layer that makes offsite monitoring economic — AI watches everything continuously and a human verifies only what it raises. It also underpins the 360° Shrink Control framework in retail, where the five zones each have their own detection profile.
Book a free site assessment.
Tell us about your site and we will come back within one business day. No obligation, no pressure.