VERBROSTECHNOLOGIES
  1. Home
  2. Products
  3. ARGUS

ARGUSVideo Analytics + LLM Insights

Turn the CCTV you already have into answers about your stores.

ARGUS is an edge-to-cloud platform that ingests RTSP and CCTV feeds and turns them into KPIs: age and gender mix, crowd count, dwell time, queues, occupancy, visits and cross-store comparisons. An LLM layer on top lets your team ask questions in plain language and create alerts just by writing them down.

Enquire about ARGUS See how it works
edge + cloud
ID 14 · F 25–34 dwell 6:12 ID 22 · M 35–44 ID 31 · F 18–24 queue 2 · wait 1:40 CAM 03 · AISLE 2 LIVE 18:42 Footfall today 1,284 Avg dwell 6m 12s +48s vs last week Occupancy 72% Billing queue 2 peak 7 at 18:10 Visitors by hour ARGUS insight Footfall peaked 6–7 pm, 31% above the weekday average. Dwell in Aisle 2 rose after the new display went up. Suggestion: open counter 3 from 6 pm on weekdays.
Runs on
Edge device on site + Verbros cloud
Video input
RTSP, CCTV and NVR streams
Output
KPIs, answers and prompt-based alerts
Best for
Retail chains, malls and branches

Your cameras already see everything. ARGUS makes it count.

Most stores and sites record hours of CCTV footage that nobody watches unless something goes wrong. The questions managers care about, like how many people came in, where they spent time and how long they waited, go unanswered or get estimated by hand.

ARGUS connects to the cameras you already have, analyses video at the edge and converts what it sees into anonymous events and metrics. Those metrics roll up by zone, store, region and hour, and an LLM layer turns them into plain-language answers, daily summaries and alerts.

What changes with ARGUS

  • Manual countsAutomatic footfall and visits for every entrance and zone
  • Staffing by gut feelRosters planned against hourly footfall and queue data
  • Unused dashboardsQuestions answered in plain language
  • Problems found laterAlerts the moment a queue or occupancy limit is crossed

How ARGUS works

Four stages take a raw video stream to a decision. Video stays on site; only events and metrics move to the cloud.

  1. Connect your cameras

    ARGUS reads standard RTSP streams from IP cameras and NVRs. There is usually no need to replace cameras; we review angles and coverage in a site survey.

  2. Analyse at the edge

    An on-site edge device decodes the streams and runs person detection, multi-object tracking and attribute models. Video becomes anonymous events: entered, left, dwelled, joined a queue.

  3. Aggregate in the cloud

    Events stream to the Verbros cloud, where they roll up into KPIs by camera, zone, store, region and hour, and can be compared across stores and periods.

  4. Ask and get alerted

    An LLM layer sits on top of the metrics. Your team asks questions, receives written summaries and creates alerts by describing them in a sentence.

Features

Nine capabilities that cover the full picture of how people use your space.

  • Footfall and visits

    Line-crossing counts at every entrance, with visits by zone and hour and the in/out balance through the day.

  • Age and gender mix

    Estimated age band and gender for each track, reported only as anonymous totals by zone and hour.

  • Dwell time

    Time spent in each zone, aisle or fixture, so you can see what holds attention and what gets walked past.

  • Queue analytics

    Queue length and estimated wait at billing counters and service desks, with peaks by hour and day.

  • Crowd count and occupancy

    Live people count per zone against the capacity you set, for safety, compliance and staffing.

  • Heatmaps and flow

    Floor heatmaps and movement paths that show hot and cold zones and how people move through the space.

  • Cross-store comparison

    Compare any KPI across stores, regions, days and campaigns, all on one consistent definition.

  • LLM insights

    Ask questions in plain language and get answers, explanations and daily summaries written for managers.

  • Prompt-based alerts

    Describe an alert in a sentence, such as a queue limit or an occupancy threshold, and ARGUS watches for it.

Ask it the way you'd ask your best analyst

Questions and alerts are written in everyday language. These are the kinds of requests ARGUS is designed for.

  • You
    Which store had the longest average billing queue last weekend?answerRanks stores by average queue length and wait, with each store's peak hour.
  • You
    Compare dwell time in the denim section between our Gomti Nagar and Hazratganj stores this month.comparePuts the same zone side by side across stores, by week and hour.
  • You
    Alert the floor manager if more than 6 people wait at billing for over 3 minutes.alertCreates a live rule on the queue metric for that counter.
  • You
    Summarise yesterday's footfall by hour and age group.summaryWrites a short summary with the hourly breakdown attached.
  • You
    Why did weekday footfall drop at Store 12 in September?explainBreaks the change down by entrance, hour and day and points to the biggest contributors.
  • You
    Every evening, tell me which zones went above 80% occupancy.digestSends a scheduled digest of zones that crossed the threshold, and for how long.

Technical specifications

ARGUS is configured per site. These are its building blocks; we size the edge hardware to your camera count and resolution.

Privacy defaults

video
analysed at the edge, on site
cloud
anonymous events and metrics
reports
aggregates by zone and hour
identity
not recognised or reported
Video sources
IP cameras and NVRs that provide RTSP streams, H.264 or H.265
Edge compute
NVIDIA Jetson-class device or an on-site NVIDIA GPU server, sized to camera count and resolution
Vision pipeline
Person detection, multi-object tracking, attribute estimation (age band, gender), zone and line logic
Model optimisation
NVIDIA TensorRT-optimised models for real-time inference at the edge
Zones and lines
Polygons and counting lines drawn per camera for entrances, aisles, fixtures and counters
Metrics
Footfall, visits, demographics, crowd count, occupancy, dwell, queue length, wait time, heatmaps
Aggregation
By camera, zone, store, region, hour, day and custom period, with cross-store comparison
Insight layer
Large language model over the metrics store for questions, explanations and written summaries
Alerts
Threshold, duration and schedule rules written in plain language
Data out
Reports and exports, with integration into your BI tools on request
Rollout
Pilot on a few cameras in one store, then expand store by store

What ARGUS changes for each team

The same metrics answer different questions for operations, marketing and leadership.

Store operations

  • Staff counters before queues build
  • Spot under-used zones and fix layouts
  • Keep occupancy within safe limits

Marketing and merchandising

  • Measure footfall and dwell before and after a campaign
  • See which displays and fixtures hold attention
  • Understand the age and gender mix by hour

Leadership

  • Compare stores and regions on the same definitions
  • Combine footfall with billing data to track conversion
  • Read a written daily summary instead of another dashboard

Use cases

Anywhere people come and go, wait or browse, and someone needs to plan around it.

  • Retail stores and chainsFootfall, conversion, staffing and layout decisions across every store.
  • Shopping mallsEntrance and floor-level footfall, crowd flow and visits by tenant.
  • Supermarkets and hypermarketsBilling queues, wait times and peak-hour staffing.
  • Banks and service branchesWaiting-area occupancy and service-desk queues by branch.
  • Showrooms and experience centresWhich products and zones draw visitors, and for how long.
  • Offices, campuses and venuesOccupancy against capacity for safety and space planning.

Questions about ARGUS

Can't find your answer here? Ask us directly.

Do we need new cameras?

Usually not. ARGUS works with IP cameras and NVRs that provide an RTSP stream. During a site survey we check camera angles, height and lighting, and suggest changes only where a metric needs them.

Is our video sent to the cloud?

No. Video is analysed on an edge device at your site. Only anonymous events and aggregated metrics go to the cloud for reporting and insights.

Does ARGUS identify individual people?

No. Tracks are anonymous and exist only to count, measure dwell and follow movement within a camera view. Age and gender are estimates used in aggregate reports.

How accurate are the counts?

Accuracy depends on camera placement, lighting and crowd density. We validate counts against manual samples during the pilot and tune zones and lines before rollout.

How do prompt-based alerts work?

You describe the condition in a sentence. ARGUS turns it into a rule on the relevant metric, such as queue length at a counter for a set duration, and notifies your team when it is met.

Can we start small?

Yes. Most teams start with a pilot on a few cameras in one store, compare the numbers with what they know, and then expand.

See ARGUS on your own store layout.

Share your store count and camera setup, and we'll propose a pilot.

Enquire about ARGUS
Emailinfo@verbros.com
Phone+91 746 000 2107
OfficeVivek Khand 2, Gomti Nagar, Lucknow
All productsCompare ARGUS, IRIS, MUSE and ATLAS Next productIRISAI Smart Mirror