AI that turns your cameras into events, triggers and analytics.
Computer vision finds the people, tables and queues on the cameras you already have. A vision-language model describes what is happening, your rules turn it into alerts, and every event adds up to analytics that Aldo acts on.
- 1Your camerasThe IP cameras you already have
- 2Computer visionFinds people, tables and queues
- 3VLMDescribes the scene in words
- 4Events and triggersYour rules decide who hears
- 5Analytics and AldoTrends, and action in Slack
Live· Location 03 · Cam 042Computer vision · 3 detections3VLM reads the scene
“Guests left Table 6 three minutes ago. Plates and glasses are still there, and no staff are near.”
4Event and trigger
Table waiting to be cleared · 3:12
Rule: over 3 minutes, alert #floor-team
5Analytics
This week, by hourTables cleared late
Aldente Vision

Cam 01 · Entrance212passers-by an hour, 38% walk in 
Cam 02 · Counter2/2staff at the counter, no queue 
Cam 04 · Dining room3:12Table 6 waiting to be cleared 
Cam 07 · Back door1departure without a clock-out
Visitor traffic
Street and inside, by hour
Occupancy
How full each room is, all day
Queues
Length and duration, counter and drive-thru
Tables
Seating time and tables to clear
Staff presence
Arrivals, departures, empty stations
Repeat visits
New and returning guests
Peak hours
Day-by-hour heatmaps
Leaderboards
Compare locations, set challenges
Set the rule once. Aldo watches for it.
Choose an event, a threshold and a person. When a queue runs long or a station sits empty, the alert goes out with the camera view.
- Queue over 6 people for 5 minutesDistrict manager
- Station unattended for 3 minutesShift lead
- Table uncleared for 3 minutesFloor team
Compare every location on the same terms
Peak-hour heatmaps, leaderboards and week-over-week trends show which locations need help and which ones to learn from.
Know your regulars
Count new and returning guests and see how often they come back, without names attached.
Put the cameras you already have to work.
See Aldente Vision in a 30-minute demo.