CCTV AI · People counting · No cloud
Turn CCTV Into Revenue & Mall Analytics
People counting and visitor analytics for malls and stores from CCTV already installed · without the cloud.
- No CCTV replacement
- No extra sensors
- Data stays on your premises
Mall intelligence
Sentral Kota Mall · Today
Visitors
18.402
Capture rate
0%
Peak hour
19:00
Traffic by hour
Traffic by entrance
- Main lobby7,270
- Parking B15,870
- Ride-hail drop-off4,750
- MRT entrance4,030
Illustration · not customer data
Trusted by malls and retailers in Indonesia
- visitors counted
- 5M+
- visitors counted
- CCTV channels in one deployment
- 300+
- CCTV channels in one deployment
- video sent to cloud
- 0 sec
- video sent to cloud
“I now have capture rate data for every tenant for renewal negotiations, not just gut feel.”
“The tech is advanced · I see store traffic without manual counting.”
Why visibel.ai
Understand what happens before checkout.
You know today's revenue. Maybe how many people entered. But how many passed and didn't come in? Which zones get skipped? Which events actually moved tenants?
CCTV has been watching in silence. Time to use it.
With computer vision, visibel.ai records what happens in between · without a second of video leaving the building.
Features
One system, from entrance to checkout
Store · Corridor A
Passer-by → walk-in
Zone × hour matrix
Which zones are alive?
Quietest: Kids corridor · busiest: Atrium, 19:00
Tenant leaderboard · Floor 1
Who draws visitors?
- 1Fashion A31%+4
- 2Cafe B27%+2
- 3Electronics C19%−1
- 4Cosmetics D12%−3
Capture rate from corridor traffic into the tenant
Saturday · occupancy vs staff
Staff at the right hour
Add 2 staff at 18:00–20:00. Cut 1 at 10:00–12:00.
Atrium event · vs normal weekend
Busy on stage, busy at checkout?
+67%
Traffic atrium
+17%
Walk-in tenant
+9%
Dwell time
Ilustrasi skenario, bukan data pelanggan.
People counting at the storefront and inside: how many pass, how many actually enter.
- Passer-by vs walk-in per door and hour
- Store and tenant capture rate
- Compare windows, hours, and days
- No manual counting
A good floor plan on paper isn't always alive in the store.
- Zone heatmaps and zone × hour matrix
- Dwell time per area
- Peak occupancy and capacity
- Basis for layout and display placement
Leasing, marcomm, and tenants finally see the same numbers.
- Storefront traffic and walk-in per tenant
- Performance rankings across floors
- Objective material for renewal
- Optional tenant portal · new revenue stream
Staff ready at the right hour beats surprise overtime.
- Peak hour per day and entrance
- Security, cleaning, and cashier schedules
- Staff exclusion for clean data
- Compare across branches
Without comparing to a normal weekend, success lives only in the caption.
- Compare campaign period vs baseline
- Measure walk-in lift, not just atrium crowds
- Reports ready for teams and tenants
- Build memory: what worked, what didn't
AI on-premise
Cloud is a choice. Not a requirement.
Instead of sending video to the cloud (heavy bandwidth, unstable connections, leak risk), analyze locally. visibel.ai brings AI to your building: edge device or on-site server connected to existing CCTV. Video is processed on-site; only numbers leave.
Cloud-based analytics
visibel.ai AI on-premise
CCTV video
Stream sent to third-party servers for analysis
Analyzed in your building. Video never leaves
Bandwidth
Continuous video upload from every camera
Only aggregate numbers are sent · lightweight
When internet fails
Analytics are disrupted too
Processing keeps running on-site
Data control
Data sits on vendor infrastructure, often abroad
Stays in your server room · easier PDP alignment
Scale
Cost and network load grow with every camera
300+ channels monitored, still processed locally
Blind spot
Assumptions are expensive.
Marcomm judges from crowds. Leasing from rent. Tenants from checkout. Everyone measures · with different rulers.
- 01
Full atrium, successful event?
Crowds on stage don't always mean busy checkout.
- 02
High passer-by means strong sales.
Busy out front doesn't always mean busy inside.
- 03
Quiet tenant, bad tenant.
Sometimes the flow is wrong, not the tenant.
- 04
Saturdays are always busy, add staff.
Schedules from gut feel lead to surprise overtime.
- 05
Promo worked, photos look packed.
Without a baseline, success lives only in the caption.
Solutions
Built to help your business
Mall Intelligence
Understand visitors, support tenants
Footfall per entrance, corridor traffic, live vs skipped zones, and capture rate per tenant · material for renewal, leasing, and layout.
Learn more
Retail Intelligence
Understand shoppers before checkout
AI for stores: passer-by vs walk-in, peak hours for staff schedules, in-store zones, and benchmarks across branches.
Learn moreBuildings, campuses, airports, and enterprise
Hundreds of cameras in one on-premise VP deployment.
How it works
Same cameras. New revenue.
No need to replace CCTV or install door people counters. visibel.ai connects to existing IP cameras, DVR, or NVR, processes video on local hardware, then sends numbers to the dashboard.
Existing CCTV
Entrances, atrium, corridors, or tenant areas become data sources. Stream access (RTSP) is enough.
On-site AI hardware
Computer vision runs on VE edge devices or VP servers in your building. Video is analyzed and stays there.
Dashboard & tenant portal
Traffic, zones, tenants, and aggregate reports. Optional tenant portal for subscribed tenants.
Getting started
Start with one problem. Prove it in 15 days.
- 1Day 1–3
Setup & baseline
- visibel.ai hardware is installed and connected to your existing CCTV
- Define priority doors, zones, or tenants
- Baseline footfall and capture rate start to form
- 2Day 4–15
Data & review
- Traffic, walk-in, and zone data collected every day
- Map busy areas and paths visitors skip
- Review results with the visibel.ai team
- 3After trial
Decide & expand
- Use data for renewal, layout, and staffing decisions
- Continue subscription if it fits your needs
- Expand to more floors, stores, or locations
Privacy
AI that reads patterns, not people.
visibel.ai analyzes patterns and behavior in aggregate · not who someone is. Video is processed on-site; only business-relevant insights leave.
- No personal identification (standard plans)
- Video stays in the building
- Only aggregate data exits
- Staff can be excluded from counts
Client voices
From teams already reading the data
Stories from mall management and store owners who started deciding with data.
★★★★★
“I now have capture rate data for every tenant for renewal negotiations · not just gut feel.”
Mall Intelligence client
★★★★★
“Turns out mall CCTV can show corridor traffic · not just security footage.”
Mall Intelligence client
★★★★★
“I can see which zones are busy and which visitors walk past without stopping.”
Mall Intelligence client
★★★★★
“Traffic data makes renewal conversations with tenants more objective.”
Mall Intelligence client
★★★★★
“Peak walk-in hour is crystal clear · security and cleaning schedules fit better.”
Mall Intelligence client
★★★★★
“The 15-day trial was enough to see our mall’s weekly traffic patterns.”
Mall Intelligence client
★★★★★
“The dashboard is simple · our leasing team gets tenant capture rate right away.”
Mall Intelligence client
★★★★★
“We can compare tenant performance across floors with the same metrics.”
Mall Intelligence client
★★★★★
“No need to replace CCTV · just connect to what you already have.”
Mall Intelligence client
★★★★★
“Zone heatmaps help us evaluate placements for new tenants.”
Mall Intelligence client
★★★★★
“The tech is advanced · I see store traffic without manual counts.”
Retail Intelligence client
★★★★★
“We can compare campaign periods against baseline mall traffic.”
Mall Intelligence client
People counting
What is people counting?
People counting is how you measure visitors passing and entering an area. Typical door sensors only count at the entrance. visibel.ai uses computer vision on existing CCTV, so beyond footfall you get walk-in, capture rate, zones, and peak hours · all processed on-premise.
visibel.ai is a people counting and visitor analytics solution powered by CCTV AI. We turn CCTV already installed in your mall or store into business data: how many pass by, how many enter, how long they stay, which zones are active, and which tenants draw visitors. Everything is processed on-premise at your site.
Many solutions send video streams to cloud servers. visibel.ai is on-premise AI: computer vision runs on hardware in your server room or office. Video never leaves the building; only aggregate numbers go to the dashboard.
Door sensors are usually installed above a single entrance and only count entries. visibel.ai uses your existing CCTV, so in addition to visitor counts you get passer-by, capture rate, zones, dwell time, and period comparisons.
Not always. We review camera type, stream quality, angles, and lighting before recommending a setup. If you can view CCTV on a phone or computer, it will likely work. Extra cameras are only suggested when coverage is genuinely insufficient.
Our hardware is plug-and-play: power on, connect to the network, then the visibel.ai team configures counting lines, zones, and corridors. Typically you only need to provide CCTV stream access (RTSP).
Not on standard plans. The platform focuses on aggregate patterns such as traffic, walk-in, peak hours, and zone activity · not who someone is. Customer loyalty features exist only on the Ultimate plan for enterprise, with separate configuration and consent.
Yes. Visits to checkout areas can indicate transactions. POS integration adds accuracy by separating checkout visits from completed sales, and adds conversion and transaction value.
Yes. APIs are available for BI, POS, CRM, or ERP. We can push data, or you can pull it.
From a single store to 300+ CCTV channels in one location. Small stores and sites use edge devices VE4–VE25; large malls and buildings use VP servers for 100+ CCTV feeds. Airports and enterprise sites use custom configurations.
Yes. Processing runs on local hardware, so visitor counting continues even when connectivity is unreliable.
The data is already there. Just read it.
Visitors already came. CCTV already saw them. Start a free trial and see the numbers yourself · no camera swap, no video leaving the building.
Trials are prioritized for sites in Jabodetabek that pass CCTV readiness assessment.