Every table, every minute — and what to do about it.
An AI box at the restaurant watches the existing CCTV, tracks every table through its service, and turns it into decisions on staffing, table mix and revenue.
📷 Restaurant's camera →
🧠 Edge AI box at the venue (~₹22K) →
☁️ Only table data ~1 KB/s →
📊 This dashboard
–
Right now
Tables occupied
–
Guests seated
–
Seat utilisation
–
guests ÷ seats at occupied tables
Longest-seated party
–
Recommended actions
Loading…
Floor
Occupancy % of tables occupied, venue time
Events
Parties arriving and leaving appear here.
What last service tells the owner
Occupancy by hour clear-view tables vs all tables
01
Uses the camera they already have
Any RTSP/IP CCTV. Draw table zones once; the box does the rest.
02
Runs on a ₹22K edge box
Raspberry Pi 5 + AI accelerator handles several cameras. No cloud GPU, no video upload.
03
Tracks the whole table cycle
Seated → dining → wrapping up → left → reset, party size and time at table.
04
Turns it into decisions
Live nudges for the floor manager; staffing, table-mix and revenue insights for the owner.