You know how many receipts your counter issued. You don’t know how many waited and left.
We count visitors on the cameras you already have and compare them with counter receipts, hour by hour, in every pharmacy. We show you when the queue crosses the threshold, when staffing drops below the minimum, and in which hours you lose customers. No new hardware, and we don’t see prescriptions. We recognize only the employees you register; patients remain silhouettes.
customerservedleft without buyingpharmacist, not counted
Centru Pharmacy · 18:00 · 42 walk-ins, 2 pharmacistsQueue 5 at the counter · alert sent in 10 sServed today 81% of walk-insPeak hours 17–19 · most staff at 14:00Morning report · sent at 07:30Chain · 12 pharmacies measured the same wayAverage wait time 2:40 · threshold 4:00
The pilot, in real data
Of 100 people who walk in, how many leave without a receipt.
31 days in a row, from 14 September to 14 October 2026. People were counted by hand on the camera images, one by one, to calibrate Vizor. The receipts come from the pharmacy’s till. Pick a pharmacy and an hour.
In production
since August 2026
Locations
two pharmacies of the same client
Client
pays a monthly subscription
Verification
manual counts and fiscal receipts
all day
30of 100 leave without a receiptall day · pharmacy A
paidno receipt
no receipt, by hourtap an hour
28%until 12:00
38%from 15:00
34%Mondays, the weakest day
19–38%best and worst day
“No receipt” means people counted minus receipts, in the same hour. It includes companions, who do not pay separately, and people who only asked something. Calculated over the hours with at least 2% of the day’s traffic: 08–17 at pharmacy A, 08–19 at pharmacy B, i.e. 97% and 99% of people. We do not name the client and do not publish how many people walk in, only percentages.
The two pharmacies have different patterns. At pharmacy A, the share leaving without a receipt grows towards the evening. At pharmacy B it is higher in the morning. That is why Vizor measures every location, every hour.
How accurate. The error measured in the pilot.
A counter that does not state its error cannot be checked. We publish ours as it came out. In the pilot, Vizor’s numbers are calibrated today with manual counts.
People counted1.5%
average daily error
Vizor’s count against the manual count, in the pilot.
Customers served1.7%
error against fiscal receipts
Customers served in the Vizor report, compared with the pharmacy’s fiscal receipts.
Calibration set33 days
counted by hand, person by person, in both pharmacies
From 14 September to 16 October 2026. The next step is automatic calibration at installation, with no manual counting.
Milestones. From the first camera to a paying client.
01
First camera connected
In a pharmacy, on the existing camera that sees the till and the queue.
July 2026
02
In production
An automatic report every day, sent to the client.
August 2026
03
Second pharmacy
For the same client. The numbers are checked against fiscal receipts and manual counts.
September 2026
04
The real purchase rate
Reports calculate customers served from the receipts of the last four weeks. The second product is in testing.
October 2026
Who is building VizorAnatolie Jivun, founder, Chișinău. Built Vizor from the first camera connected in a pharmacy: the installation, the models that count and the daily reports the client receives.
What Vizor sees in a pharmacy
Three things the cash register doesn’t know.
The cash register counts receipts. Vizor counts the people who came in, how long they waited at the counter, and how many left before getting there. The difference is where shift and counter decisions come from.
0
Walk-ins vs served
How many come in, how many reach the counter.
By hour and by day, with pharmacists told apart from customers by behavior. The difference is the number nobody has today.
queue 5 · counter · open the second one
Queue at the counter
How many wait, and for how long.
Once the threshold is crossed, the shift supervisor gets an alert on their phone. In a pharmacy, a four-minute wait means customers leaving for the pharmacy next door.
Staff vs customers
The shift is built for 14:00. Customers come at 18:00.
The shift schedule is compared with actual traffic, hour by hour. Usually you’re not short of people; you’re short of them at the right hour.
Theft and shrinkage
We see when goods leave without stopping at the counter.
In the self-service area — dermocosmetics, vitamins, children’s products — the same cameras watch for three theft patterns. When one occurs, an alert goes out with the exact time and the exact camera. Checking takes ten seconds instead of an hour of searching through recordings.
time · camera
Exit without stopping at the counter
Someone takes products from the self-service area and leaves without stopping at the counter. The alert includes the camera and the exact second.
18:42:07 · camera 2 · exit bypassing the counter
expensive products
Unusually long dwell
Someone lingers at the shelves with expensive products, outside the usual pattern for that hour. A pharmacist stepping out from behind the counter usually settles it.
16:15:40 · camera 1 · 6 min at dermocosmetics
gesture, not product
Item hidden in a pocket or bag
The system tracks the gesture, not the product: the hand moving from the shelf to a pocket, bag, or coat. It’s enabled only on the shelves you choose. If the person then leaves without stopping at the counter, the alert gets top priority.
11:03:22 · camera 1 · hand-to-bag gesture
Pharmacy dashboard
The Vizor dashboard, as it looks in the real installation.
The dashboard layout from the real installation. Switch pharmacies and go through Dashboard, Alerts, Analytics, and Reports. The figures are illustrative; in the installation they come from the pharmacy’s camera and the counter receipts.
app.vizor.md · pharmacies · dashboard
V
Vizor
On the phone
The same numbers, with the shift supervisor.
The mobile app is for whoever makes the call in the next minute: the shift supervisor, the pharmacy manager, the head pharmacist. No tables, just three things that matter right now.
The alert reaches the lock screen with the pharmacy, the camera, and the exact minute. Checking takes ten seconds.
Walk-ins, served customers, and the wait at the counter for the current day, in the selected pharmacy.
The evening report takes a minute to read, right in the app or in Telegram.
The day’s numbers reach your phone, without customer names.
Pharmacy reports
This is the report that arrives every evening.
Not presentation charts: the document the pharmacy manager receives in Telegram at 20:30, with the day’s numbers and two masked frames. The format is the same as in our real installation; the figures here are illustrative.
Daily report
Centru PharmacyThursday, September 4, 2026 auto-generated at 20:30 · cam 1
Walk-ins298+6% vs last Thursday
Served at the counter24181% of walk-ins
Left unserved5719% · 31 of them 17:00–19:00
Peak hours18:0042 walk-ins · 2 pharmacists
Wait at the counter2:40max 5:10 at 18:04
First · last movement08:02 · 20:51open 08:00–21:00
By hour · walk-ins and served
Hour
Walk-ins
Served
Rate
Max queue
08
8 / 7
88%
1
09
14 / 13
93%
1
10
19 / 17
89%
2
11
22 / 20
91%
2
12
26 / 23
88%
2
13
28 / 25
89%
3
14
25 / 23
92%
2
15
23 / 20
87%
2
16
27 / 23
85%
3
17
34 / 24
71%
5 · 17:48
18
42 / 29
69%
5 · 18:04
19
28 / 24
86%
3
20
9 / 9
100%
1
Moments of the day · masked frames
CAM 1 · 2026-09-04 First movement · 08:02faces maskedCAM 1 · 2026-09-04 Day’s peak · 18:04 · queue 5faces masked
For tomorrow
1Second counter open between 17:00 and 19:00. Yesterday, 31 people left unserved in this window.
2Second shift starts at 12:00 instead of 10:00: in the morning, the service rate is above 88% with a single pharmacist.
3Tuesday, 10–12: manual count for calibration. We compare the numbers together.
Vizor · Centru Pharmacybot · daily report and alerts
Thursday, September 4
Queue of 5 at the counterCentru Pharmacy · cam 1 · 17:48. One pharmacist at the counter, 5 people waiting. Open the second counter.17:48
Queue back below threshold18:19 · wait 1:50. The previous alert closed itself.18:19
Daily report · September 4Walk-ins 298 · served 241 (81%) · left unserved 57 Peak 18:00, 42 walk-ins · max queue 5 at 18:04 First movement 08:02 · last 20:51 Tomorrow: second counter 17–19.
We have one pharmacy measured daily, with a parallel manual count. The scenarios below are built on the patterns measured there and use modeled figures, labeled as such.
Real case · in productionpharmacy · Moldova
The first pharmacy measured daily, with a parallel manual count.
Situation
A pharmacy with one counter, existing cameras, and no numbers on how many people come in or how many leave unserved.
What we do
We count walk-ins and served customers by hour, separate pharmacists from customers, and send the daily report every evening, with masked images. A few hours a week are counted manually, for calibration.
What came of it
The method we propose for the pilot: numbers checked together, not promised by the vendor. The peak hour and the hour with the weakest service rate don’t match what the team thought.
Aug 2026in production; automatic report every evening
walk-ins by hour · a typical dayserved
Modeled scenarioPharmacy chain · 12 locations
18:00 brought the most people and the fewest receipts.
Problem
Evening sales weren’t growing, even though evening traffic was the highest of the day.
What it showed
The queue exceeded 5 people in 6 of the 12 pharmacies between 17:00 and 19:00. The service rate dropped by 11 points at exactly that time.
Decision
Second counter opened for 17–19, second shift moved by an hour. Zero new hires.
A single pharmacist after 18:00. A queue of six at 18:30.
Problem
Reviews complaining about the wait, customers leaving for the pharmacy across the street, with no way of knowing how often.
What it showed
Staffing peaked at 14:00, when 19 people an hour came in. At 18:00 it was 34, with a single pharmacist. Average wait: 6 minutes.
Decision
Shift overlap moved from 13–15 to 17–20. Same number of paid hours.
−55%wait time at peak hour, within 2 weeks
pharmacists vs customers · by hourbefore
The modeled scenarios use the patterns measured in that pharmacy and effect sizes from the retail analytics literature. During the pilot, we measure your own numbers together with you, alongside a parallel manual count.
How it works in a pharmacy
One day of installation. No need to close.
01
Your current cameras
Usually one camera that sees the door and the counter is enough. Where it isn’t, we add a standard one.
zero investment
02
Automated analysis
The video stream is analyzed continuously. Pharmacists are told apart from customers by behavior and position.
own method
03
Numbers, not identities
Images are processed on the Vizor server; the report gets numbers. We don’t read prescriptions and don’t link receipts to people.
data minimization
04
Link to the POS
The counts are matched with the counter receipts, which gives the real service rate.
service rate
Privacy in the pharmacy
We don’t identify customers. We don’t read prescriptions.
The system tracks silhouettes from frame to frame and produces numbers: how many came in, how many were served, how long they waited. Images are processed on the Vizor server. We provide the customer privacy notice and the compliance documentation. Individual monitoring applies only to employees you explicitly register, with prior notice to them and the documented legal basis required by Law No. 195/2024.
Pilot for pharmacy chains
Three pharmacies. Three months. Numbers checked together.
One high-traffic, one small, and one neighborhood pharmacy, so the result says something about the chain. In the third week, we count manually in parallel, during hours you choose. You can stop the pilot at any time, with no penalty.
We don’t ask you for
New cameras
We work on your existing cameras. Where coverage falls short, we fill the gap.
We don’t ask you for
Your pharmacists’ time
Installation takes one day and doesn’t interrupt operations. No integration required to start.
We don’t ask you for
Long contract
Three months. You can stop at any time. The comparison report is yours to keep.
We don’t ask you for
Patient data
We don’t need prescriptions or product data. Only the video stream and, if you want the service rate, the number of receipts per hour.
Calculator · what a higher service rate is worth
In a typical pharmacy, the service rate drops by more than ten points at peak hour compared with the rest of the day. Here we calculate conservatively, with the figure you choose. An estimate, not a promise.
Extra sales per month180,000 MDL
extra receipts / month1,125
per year2,160,000 MDL
Calculation: visitors × days × service-rate points × average receipt × pharmacies.
Does the system see what customers buy, or their prescriptions?
No. The system tracks silhouettes and produces numbers: how many came in, how many were served, how long they waited. It doesn’t read prescriptions or identify products. It recognizes faces only for the employees you register.
How does it tell pharmacists from customers?
By behavior and position: who stands behind the counter, who comes in and goes out through the door. Pharmacists aren’t counted as visitors. If you register them, they’re recognized by face and can be reported on individually.
Is one camera enough for a pharmacy?
Yes, if it sees the door and the counter. That’s exactly the setup in the pharmacy where we already work. Where coverage falls short, we add a standard camera.
How long does installation take?
One day, without closing. No integration required to start; the link to counter receipts is set up once the traffic numbers are validated.
How accurate are the numbers?
In the third week of the pilot, we count manually in parallel, during hours you choose, and check the numbers together. In the pharmacy we measure daily, manual calibration is part of the weekly routine.
How do we start?
With a three-month paid pilot in three pharmacies, which you can stop at any time, with no penalty. Before that, 30 minutes and a look at the camera in one pharmacy.
Contact
Show us a pharmacy. We’ll tell you what we can measure.
We don’t ask for an hour-long meeting. You pick a pharmacy, we look at the existing camera, and we reply within 24 hours, in writing, with what can be measured there.
What happens next
Three steps, no pressure.
1We reply within 24 hours, in writing, with what can be measured at your location.
2Thirty minutes on a single location: together we look at what your existing cameras cover.
3If it makes sense for you, we propose a pilot on three locations. You can stop it at any time, with no penalty.