+12.4%
Entries today128Directional crossings today Coderoffice Vision · Interactive prototype
Retail camera analytics your team can actually use.
Turn compatible camera feeds into clear operational signals: footfall, entry and exit counts, occupancy, queues and zone activity — without asking a manager to watch hours of footage.
Anonymous by defaultNo customer face recognition in this demoEdge-first processingAnalyse video near the cameraReal-time capableMeasured on the prototype test stream
Demo Store 01 · Front entrance Live
DEMO FEED · --:--:--13 FPS · 634×360 · AI overlay
Person 0.96P-142
Person 0.91P-143
Person 0.88P-144
ENTRY LINE 01INOUT
Camera 01 · Bi-directional line crossing3 people in frame · anonymous tracking
Entries128+12.4%
Inside now7Normal
AI throughput49.6FPS test
Live event↗ Visitor enteredFront entrance · just now
PrivacyAnonymous analyticsOperational signals, not identities
Working prototype benchmark49.6 FPSprocessing19.14 msavg. inference3.8×source rate489 MBpeak process RAM
01 · Product experience
Give the client something they can click, not just something they can read.
This interactive concept uses simulated business metrics around a real working computer-vision prototype. Try the navigation, camera views, date range, alerts, zone map, export and demo event controls.
Interactive demo. Operational numbers shown inside the product UI are simulated for demonstration. Prototype performance figures are measured from the local test run.
WorkspaceDemo Store 01 · Adelaide
Demo Store 01 · Adelaide
Store overview
Live operational signals from the interactive product prototype.
Live · --:--:--
+9.8%
Exits today121Directional crossings todayNormal
Currently inside7Estimated live occupancy−1.2m
Average dwell14m 08sDemo session estimateVision agentOnlineCameras3 / 3 onlineAvg. inference19.14 msPrivacy modeAnonymous
Front entranceCamera 01 · Bi-directional line crossing
DEMO FEED · --:--:--13 FPS · 634×360 · AI overlay
ENTRY LINE 01INOUT
Camera 01 · Bi-directional line crossing3 people in frame · anonymous tracking
Live eventsNewest activity first
MonitoringAnonymous track inspectorClick a green detection box to inspect the signal
No track selected
Select a bounding box in the camera view. The demo exposes operational tracking data, not a person’s identity.
Directional trafficHourly visitor movement
Entries Exits
OccupancyEstimated people currently inside
Limit 307people
Normal activity
Current occupancy is 23% of the configured demo limit.
Updated from entry / exit eventsTry the demoTrigger an event and watch the dashboard update.
02 · What the pilot can measure
Start with four signals a store manager understands immediately.
A focused first release is easier to validate than a dashboard full of speculative AI features. The system can expand after the entrance-counting workflow proves useful.
01Footfall
Entry & exit counting
Convert directional line crossings into hourly and daily visitor traffic.
02Operations
Live occupancy
Estimate how many people are inside a monitored area using entry and exit events.
03Service
Queue awareness
Monitor checkout zones and trigger an event when queue size exceeds a configured threshold.
04Layout
Zone activity
Compare anonymous movement and dwell patterns across selected areas of the shop.
03 · Edge-to-cloud architecture
Keep the video workload local. Send useful events to the business platform.
The vision agent handles detection and tracking near the camera. The web platform receives lightweight events and summary data for dashboards, alerts and reporting.
04 · Prototype evidence
Measured performance, separated clearly from simulated product data.
These figures came from the local prototype test described earlier. They show technical feasibility for that configuration and should not be presented as a guarantee for every camera, PC or deployment.
Smart Shop Vision · Prototype
Start with one camera, one entrance and one business question.
A small pilot can validate camera placement, counting quality and whether the resulting data is useful before you invest in a wider rollout.