Do I need new cameras?

No. ZeoVision works with the IP cameras you already have. The survey checks the angles first — a camera pointed at the ceiling can't count anyone, and we will tell you that before you commit.

Will it work with my camera brand?

Almost certainly. Hikvision, Dahua, Axis, Uniview and most other IP cameras work. If yours is unusual, say so on the form and we will check before anything is agreed.

Do you recognise faces?

No. There is no facial recognition and no identity tracking in the product. ZeoVision counts a person as a person and has no way to tell you who they were.

Where is the video processed and stored?

Video is processed on your own site and stays there. Only counts and events — numbers and timestamps — leave the building, and those sit in your own country in twelve countries, the UAE and the Gulf included.

Can my staff be monitored individually?

No, and we want to be straight about the limits. ZeoVision reports coverage by zone: how many people are on the floor and where against where customers are. It does not identify who is who, so it cannot produce a performance report on a named employee, and we will not build one.

How long does setup take?

The connection takes a day or two. Then ZeoVision watches for two weeks before it recommends anything, so you get the first real report about three weeks after we start.

What does the beta cost?

We are onboarding a small number of founding partners, at a discounted rate for early partners. We are not publishing a price list while the product is in private beta, because the right number depends on your sites and cameras. Fill in the form and we will give you a straight figure for your setup.

Illustration

ZeoVision · Al Barsha branch

Visitors today

1,284

+12% vs last Thursday

Checkout queue

Checkout · 8 waiting · threshold 6

2 over threshold for 14 min

Zone heatmap · today

Share of foot traffic

This week’s recommendation

Add a second cashier on Thursdays, 6–8pm.

Queue passed 6 people for 94 minutes last Thursday with one till open.

Mark as doneNot useful

See what's happening in your business — as it happens.

ZeoVision connects to the cameras you already have and turns them into a live picture of your space: how many people come in, where they go, where queues build up, and whether you have enough staff on the floor.

Join the beta Speak to a human

Works with your existing IP cameras. No new hardware. Video never leaves your site.

Your cameras record everything. They tell you nothing.

You have months of footage and still can't say how many people walked in last Tuesday.
The queue builds at peak and people put the basket down and leave. Nobody counts them.
The rota is built from habit and last month's guess, not from when customers actually arrive.
Half the shop is busy and half of it is dead, and you only find out when stock doesn't move.
A missing helmet or a fall gets noticed when someone happens to scroll back through the recording.

Nothing to buy. Nothing to install on the wall.

Connect the cameras you already have

Hikvision, Dahua and the other major brands. We connect to your existing IP cameras and recorder. You buy no hardware and nobody runs cable through your ceiling.

It counts, measures and maps on its own

People in and out by hour and entrance, queue length at the counter, how long people stop in each zone, where staff are standing against where customers are. All of it on site. Only the numbers leave the building.

Read the recommendations, not the dashboard

Each week you get 3 to 5 sentences with the evidence underneath. "Add a second cashier Thursdays 6–8pm." You mark each one done or not useful.

Eighteen things your cameras could already be telling you.

Traffic & flow

Foot traffic counting

Trigger: People walk in
What happens: They are counted by hour and day, per entrance
Outcome: You know your real peak, not the one you assumed.

Zone heatmaps

Trigger: Customers move around your space
What happens: A heatmap builds by zone and hour
Outcome: You learn which shelf nobody ever reaches.

Dwell time

Trigger: A customer stops in front of a display
What happens: How long they stay is measured
Outcome: Long dwell and no sales means the display is wrong, not the product.

Customer flow paths

Trigger: People walk from the door to the till
What happens: The routes they take get mapped
Outcome: Lay the place out the way people actually walk.

Entry/exit conversion

Trigger: Footfall is compared to transactions
What happens: You get a conversion rate by hour and by day
Outcome: "Busy day, bad day" turns into a number you can work on.

Queues & staffing

Queue length monitoring

Trigger: A line forms at checkout
What happens: Its length is measured continuously
Outcome: You see it building, instead of hearing about it after customers leave.

Queue alerts

Trigger: The queue passes the number you set
What happens: An alert lands on WhatsApp
Outcome: Someone opens a second till now, not in ten minutes.

Walkout detection

Trigger: Someone joins a queue and leaves before reaching the front
What happens: It is counted as a walkout
Outcome: You can finally put a number on what waiting costs you.

Staff coverage

Trigger: Customers crowd one end and staff stand at the other
What happens: The gap between where staff are and where customers are is flagged
Outcome: Put your people where the customers are.

Staffing vs demand

Trigger: Footfall by hour is compared to your rota
What happens: The hours you are over- and understaffed are listed
Outcome: Build the schedule from demand instead of habit.

Occupancy counts

Trigger: A zone fills toward its capacity
What happens: A live count runs against the limit you set, with alerts
Outcome: Compliance and comfort, without anyone holding a clicker.

Safety

PPE compliance

Trigger: A worker enters a restricted area without a helmet, gloves, mask, safety shoes or uniform
What happens: A timestamped event is logged and flagged
Outcome: Correct it this shift, not in the quarterly audit.

Restricted-area entry

Trigger: Someone walks into a zone they shouldn't be in
What happens: An alert goes out straight away
Outcome: You deal with it before it becomes an incident.

Fall detection

Trigger: A person falls
What happens: An alert goes out immediately
Outcome: Help arrives in seconds, not when someone happens to walk past.

Reporting

Dead zone detection

Trigger: A corner gets a fraction of the traffic the rest does
What happens: It is flagged with its share of footfall
Outcome: Move the promo display there and watch the share change.

Multi-branch comparison

Trigger: The same metrics run in every branch
What happens: Branches are ranked side by side on footfall, queues and conversion
Outcome: You see which branch is genuinely underperforming and which is just smaller.

Weekly recommendations report

Trigger: A week of observation finishes
What happens: 3 to 5 plain-language actions are written up, each with the evidence under it
Outcome: A list of decisions, not a dashboard to study.

Coming soon

Energy control (coming soon)

Soon

Trigger: A zone stays empty for several minutes
What happens: Lights and AC will step down on their own
Outcome: You stop cooling and lighting an empty room. This one is on the roadmap, not live yet.

What a week of watching actually tells you.

Example report

Checkout queue passed 6 people on Thursdays 6–8pm with one cashier. Add a second cashier in that window.
The back-left zone gets 4% of foot traffic. Consider moving your promo display there.
3 workers entered the warehouse floor without helmets this week — all during the morning shift.

What is the queue costing you?

Put in what you know about your own shop. The numbers below move as you type. They are an estimate, and the caveat under them means it.

Estimated walkouts per week: 3–5
Estimated recoverable revenue per month: AED 650–1,050
Estimated staffing hours misallocated per week: 16–27

An estimate from your own inputs and published industry benchmarks, not a measurement of your business. A pilot replaces these numbers with real ones.

Find the one that sounds like your floor.

Find out which half of your shop nobody walks into.

We count people. We don't identify them.

No facial recognition and no identity tracking. ZeoVision counts a person as a person. It does not know who they are and it cannot tell you.

Video is processed on site. Only counts and events leave the building — numbers, not footage.

Your data stays in your country, in twelve countries including the UAE, Saudi Arabia, Qatar and Kuwait.

Read our privacy page

Where the other two options leave you.

Compare on	Enterprise video AI	Traditional CCTV	ZeoVision
Hardware	New cameras required	Already installed	Your existing cameras
Output	Dashboards and analysts	Recordings nobody watches	Plain-language recommendations
Who operates it	A safety or operations department	Nobody	The owner, on a phone
Video leaves the site	Usually	No	No — counts and events only
Priced for	Enterprise	A one-off purchase	Small businesses and multi-branch

Our targets — not guarantees.

Reduce queue wait times 15–25%
90%+ people-counting accuracy
Staffing matched to real demand

What happens after you sign up

1
Site survey
We look at the cameras you have, what they see and what they miss. If your angles can't support counting, we say so then — before you spend anything.
2
Camera connection and zone setup
We connect to your existing cameras and draw your zones with you: entrance, checkout, the aisle you care about, the area staff shouldn't enter.
3
Two weeks of baseline observation
ZeoVision watches and says nothing. Two weeks gives us your real pattern, so the first recommendation is based on your shop and not on an average one.
4
First recommendations report
3 to 5 actions with the evidence under each. We go through it with you, and you tell us which ones were worth the paper.
5
Ongoing support
A report every week, alerts on WhatsApp as things happen, and a person you can message when a number looks wrong.