Guest recognition
Every returning face becomes a profile: visit count, dwell time, favourite items and lifetime value — built without a single loyalty card.
Visual intelligence for restaurants & cafés
Your cameras already see everything. Now they understand it.
One ordinary webcam is enough to recognise returning guests, measure how long they wait, see what your baristas are actually doing, and fire the right promo before the order is placed.
- < 1s
- Face match latency, on-device
- 128-d
- Signature per enrolled face
- 6
- Behaviour signals per frame
- 0
- Seconds of video stored
Six modules, one camera feed
Built for F&B operations first: loss prevention, guest experience, team productivity and outlet analytics in a single console.
Staff monitoring
Face-based clock-in, active vs idle minutes, orders served and service speed per barista. Attendance fraud disappears.
Traffic & tables
Hourly footfall, queue pressure, dwell distribution and table turn — the numbers that decide staffing and layout.
Promo engine
Rules match a recognised guest to the right offer and greeting the moment they walk in, then measure redemption.
Behaviour AI
AI vision reads the scene: who is staff and who is a guest, mood, activity and how many people are queueing.
Privacy by design
Detection runs in the browser. Only mathematical face signatures and behaviour metadata are stored — never raw video.
From a face to a decision
- 01
Detect
Browser-side detection finds every face in frame, no cloud upload.
- 02
Match
The 128-d signature is compared to enrolled guests and staff.
- 03
Understand
AI vision reads role, mood, activity and queue length.
- 04
Act
Greeting, promo, staffing alert or loss-prevention flag fires.
Start with one camera at one outlet
Enrol a few faces, walk in front of the lens and watch the profile, behaviour stream and promo trigger build themselves in real time.
Open the console