Live LUMEN · Retail Intelligence

Every shopper, aisle and sale —understood in real time.

Who your best customers are, why shoppers leave empty-handed, which aisles earn. LUMEN reads it off the cameras you already own — on-premise, live.

Request a pilot See what you would recoverSee the plans
Video never leaves the store Works with existing CCTV
+23%
Average basket
3×
Per $1 of fee
≈2 mo
Full payback
3 days
To go live

A model, not a client result: modelled on 3 branches · 980 receipts a day · a $14 receipt — put in your own numbers

LiveReal recording · store entrance cameraNot a mock-up: the running system's screen

Scroll to explore
The running productNot a slide — open the dashboard here
A free 3-day pilotOne branch · your own cameras · the report is yours
On-premiseVideo never leaves the building · UZ law
Bayonic is already in productionMaxWay · 22 locations · SERVO
The invisible truth

Same revenue. Completely different branch.

Two branches of the same chain closed the month on $372K. On the report they are identical. On camera they are not.

What the report seesDemo data
Branch A$372K26 040 receipts
Branch B$372K26 040 receipts
What the camera sees
Branch A42 000 visitsconversion 62%
Branch B31 000 visitsconversion 84%
BuyersLeft without buyingextra walk-outs at branch A
11 000more customers leave branch A empty-handed every month
If branch A sold at B’s rate, per month+$132K+9 240 receipts × $14
Request a pilot

An illustrative model with internally consistent arithmetic — both branches ring up ~26,000 receipts at the same $14 average. We measure your real numbers in the free three-day pilot.

Why a plain counter is not enough

A counter tells you how many. LUMEN tells you who.

One month, one door — and two different answers.

Demo data
People counter
  1. Walked in126,000
  2. Bought?the counter can't see this
  3. Came back within 30 days?the counter can't see this
Who came in — unknown
?

The count ends here.

LUMEN · face recognition
  1. Walked in126,000
  2. Bought70% conversion88,200
  3. Came back within 30 days31% of buyers27,342
The same month — by who was in it
58%24%12%6%
New 73,080Bronze 30,240Silver 15,120Gold 7,560
The owner's 7 questionsyespartlyno1/77/7
How manyStaff removedNew or regularWhich aisle is deadTier at checkoutBefore they goNext weekPeople counterLUMEN

"Walked in" and "Bought" come from your calculator inputs; the rest is illustrative and measured during the pilot.

Your store

Where is your store losing money?

Nine questions, each answered by the cameras you already have. Pick one: one branch, one day, and every number shows its working.

Shoppers

Who came in today, and who comes at which hour?

Chilonzor · todayDemo data
The figures below come from this camera: everyone who walks in is counted, a recognised customer shows their tier, and staff are taken out.Recording from another store · figures are a sample day
Shoppers today1,284counter 1,430 · staff −146
Returning shoppers53942% of shoppers today
Gold shoppers778 or more visits
Women56%most active age: 25–34
Tier · today's shoppers
  • New74558%1 visits
  • Bronze30824%2–4 visits
  • Silver15412%5–7 visits
  • Gold776%8+ visits
Age × hour · who comes when

Mornings belong to shoppers aged 45–54, evenings to 25–34. 56% of shoppers are women.

Decision

Time promotions and staff by the hour: older shoppers in the morning, younger ones in the evening. Gold shoppers show up at the till with their tier.

01Customer Recognition07Demographics08Customer Segments

In a free three-day pilot we pull these nine answers from your own store’s cameras.

Request a pilot
The step after the console

Upload the report — get the answer

The console exports the day as spreadsheets. Ask a question — LUMEN's analyst reads those files and answers: one figure, the evidence under it, and one next step.

LUMENDEMO

Why did conversion drop yesterday?

Because the queue backed up at the tills in the evening — 40 shoppers left without a receipt.

  • 57%evening conversion — the day ran at 67%
  • 17shoppers abandoned their basket at the till
  • 4/5tills were open at the time
What caused it, and what the data rules out

The day ran at 67% conversion, but between 18:00 and 21:00 it fell to 57% — 10 points. 396 people came in during those hours and 40 receipts were never rung.

Mechanism. The queue crossed 12 people at 18:30 and the reserve till opened 3 minutes later. The moment the wait passed the 3:00 threshold, people started leaving baskets behind — 17 of them. The lost receipts are not in the traffic, they are in checkout capacity.

  • The data rules out
  • "Fewer people came" — the opposite: 19:00 was the busiest hour of the day.
  • "The cashiers were slow" — speed was normal, about 51 receipts an hour; the problem is that 4 of 5 tills were working.

Conclusion. The evening dip is not price and not assortment, it is capacity: past the threshold, every extra minute of queue costs a receipt.

How it is measured. Watch 18:00–21:00 conversion for 14 days: it should climb from 57% towards the day's 67%; the control figure is the number of waits longer than 3:00.

Next step

Set the rule: queue over 12 people calls the reserve till automatically — the alert goes straight to the shift lead's phone, reaction under 3 minutes.

  • WhoShift lead · reserve cashier
  • WhenFrom tomorrow, 18:00–21:00
  • Expected effectAbout 40 receipts a day come back
Files read:tashriflar.xlsxcheklar.xlsxnavbat_15min.xlsx

The four that return the most

The same cameras answer these too.

Share of shoppers reaching each zone03Customer JourneyWhere shoppers turn back+$774/mo from one layout change
People in the queue · now and in 10 minutes06Smart CheckoutThe queue, ten minutes ahead+$493/mo in freed hours and stopped shrinkageSee it live See it live
Monthly revenue · per m²02Store HeatmapWhat each square metre earns+$370/mo recovered from dead zones
Share of customers, and the revenue they bring10Loyalty & LTVYour most valuable customers73%of revenue comes from 28% of customers
Average basket · anonymous vs recognised01Customer RecognitionFirst visit or tenth+23%average basket of a recognised customer
Conversion · by shift04Employee KPIWhich shift actually sells+18%once the weak shift is brought up to the strong one
Demand · forecast vs actual, 4 weeks05AI PredictionNext week's demand94%four-week demand forecast accuracy
Visits · by age group07DemographicsWho your floor really serves+$440/mo — from advertising to the real audience
Retention · left alone vs acted on08Customer SegmentsWho is about to leave31%share of at-risk customers won back
Service score · by hour09Service & CXWhen service sags+$334/mo — from fixing the hours service drops in

The money figures come from the calculator in the next section and follow whatever you set there. put in your own numbers

Your numbers

How much would you recover in a year?

Set five sliders to your own numbers. We show money after margin, not revenue.

Currency

Your till cannot see it: this means ~1 400 visitors a day. We measure the real figure in the free pilot.

Grocery ~22% · electronics ~14% · apparel ~50%

Current monthly revenue$1.3M
What this model assumes
Up to +0.64 pp conversion — less, the less headroom you have left
0.17% of revenue — bigger baskets and shelves that are not empty
Both effects touch the same shoppers — 20% taken back off the total
Shrinkage recovered: 0.15% of revenue — already profit, no margin applied
Checkout and queue hours: $333 per branch a month — also net
Month 1 at 30%, month 2 at 65%, 100% from month 3 — the rollout curve
Platform: $600 per branch a month · One-time rollout: $490 per branch · No hardware
Net profit in the first year
$35K
$3.5K · net per month · from month 3
+$58K back over the year−$23K subscription and setup
Even in the cautious case:$6K
You pay a month$1.8Kof your turnover 0.14%You get a month$5.3K3× back every month
What each of your months looks likebase 2 monthscautious 5 months
0+$35Kinstall3 months6 months9 months12 months

The first two months are install and tuning. From the third the system runs at full strength.

After LUMEN is installed — the working+$5.3K a month
Today70%
With LUMEN70.6%

Extra buyers per month: +812 people

Added to revenue
Revenue from the new buyers+$12K
Bigger baskets, shelves not empty+$2.1K
Overlap between the two (−20%)−$2.7K
Additional revenue+$11K
Gross margin 22%+$2.4K
Money that never was revenue — profit directly
Shrinkage recovered+$1.9K
Checkout and queue hours+$1K
Net added every month+$5.3K
Platform fee−$1.8K
Rollout — one time−$1.5K
Where the money comes fromTotal a month: $5.3K
Lost shoppers and basket45.4% $2.4K
Shrinkage and time54.6% $2.9K

Even if the shopper estimate proves wrong, shrinkage and time alone cover the monthly fee 1.6×.

Pilot

Get the numbers you just set, in writing.

$3.5KNet per month
3×Per $1 of fee

3 stores · 980 receipts/day · ø$14

The figures come from the table above: each of the eight leaks is priced separately, and the platform fee is subtracted. See the arithmetic

We usually reply within one business day. No spam.

Where your data lives

Nothing leaves your building.

The cameras, the server and the data all stay inside your own store.

Your store12 camerasRTSP · LANGPUPCruns offline, 24/7numbers onlyYour dashboardNothing crosses this lineVendor cloud, abroadRaw video
1The frame is read2It becomes a numberThe image is deleted0 photos stored
Point by pointCloud vendorsLUMEN
Where AI runsVendor cloud, abroadYour server, in your store
Where video goesUploaded or streamed outNever leaves the building
Where numbers are storedForeign data centresInside Uzbekistan, yours
Personal Data Law of the Republic of UzbekistanAnonymous mode availableFull audit trail
Pricing is open — no hidden feesA monthly fee per branch. No hardware to buy.0upfront3 dayspilot, freemonthlybilling · stop any timeSee the plans
Implementation Plan

Live in three days. One workshop of yours.

Every day in two lines: what that day takes from you, and what you hold by the end of it.

  • 3 daysfrom audit to live
  • 1manager workshop — your time
  • 0new cameras — yours are used
  1. Day 1

    Audit and install

    From you
    Access to cameras and network
    You have
    A camera-coverage map and your GPU server in place
  2. Day 2

    Integrate and tune

    From you
    Nothing
    You have
    Existing cameras connected, POS / 1C in sync, models tuned to your store
  3. Day 3

    Train and go live

    From you
    A manager workshop
    You have
    Dashboards per role and pilot sign-off. From that day, in production.
After go-live
  • 24/7 monitoring from Tashkent
  • A weekly success manager
  • A report every morning at 09:00
Let's build it together

Request your free pilot

Three days, one branch, our pilot kit, zero cost. You end up with a real visitor and conversion report for your own floor — whether or not you go ahead.

  • No hardware to buy
  • Runs on your CCTV
  • Report is yours to keep
Request a pilotSee the plans
+998 87 355 88 55

We use your details only to arrange the audit. No newsletters, no third parties.