AI CCTV intelligence & operational analytics
The cameras you already have, counting visitors, not faces.
An AI solution for the IP cameras and recorders a business already runs: counting walk-ins, estimating waiting-room occupancy and waits, warning the desk before an area overflows, sending a plain report each day and, when a manager asks, finding a team member on shift. Video is analysed on site and never leaves.
We build it for your site as a project, starting with a pilot at one location. It is not a ready-made product you can switch on, and the moving pictures on this page are simulations.
Video in. Numbers out.
The camera sees a room. ResortOS would keep only what a manager needs from it: how many came in, how many are waiting, roughly for how long, and when it is getting too full.
- Visitors are counted as moving shapes. Nobody is recognised by face.
- A count is made when someone crosses a line at the door, in either direction.
- When the waiting area nears the level you set, the desk is told.
What ResortOS would keep
- Entries today
- 146
- Exits today
- 133
- Waiting now
- 8of 12 seats
- Approximate wait
- about 19 min
Flowing normally
No video, no faces, no names: only these numbers.
Part of the solution, optional
Find a team member, when it matters.
A manager types a name. ResortOS checks the person is on shift, searches the cameras in work areas for the badge on their uniform, and shows where they were last seen, usually within seconds. The team member is told that someone is looking for them.
- A guest needs urgent help: find the nearest housekeeper or technician
- An emergency or a drill: see where each person on shift was last seen
- A delivery at the gate: find the storekeeper without phoning round
Find a team member
- On shift?Waiting for a name
- Work-area cameras searched for the badgeNot started
Where
…
Camera 5, seen just now
Anu has been told on the staff app that the duty manager is looking. This search is in the audit log.
Located on request. Never followed.
Why a badge and not a face: face recognition needs a database of everyone's face, and its mistakes land on real people. A badge is visible, belongs to the job, and comes off at the end of the shift.
Before it is switched on, we would agree a written notice for your staff and a policy with you, with India's Digital Personal Data Protection Act, 2023 in mind. ResortOS does not give legal advice.
- Find a team member by name, on request
- Recognition by the staff badge on the uniform, never by face
- On-shift check against ResortOS attendance before any search
- Last-seen area and camera, usually within seconds
- A message to the team member on the staff app
- Every search in the audit log, visible to the person searched for
Badge, not face
A visible staff badge on the uniform is what the cameras recognise. It comes off at the end of the shift; a face never does.
On shift only
Only staff clocked in on ResortOS can be searched. Once someone clocks out, nobody can find them.
On request, not a trail
The answer is the last place seen, from sightings kept for 15 minutes. There is no movement history to look back through.
Work areas only
Guest rooms, changing rooms, restrooms and staff quarters have no cameras in the design and are never searched.
The person is told
The team member gets a message on the staff app saying who is looking for them, so nothing happens behind their back.
Every search on record
Who searched, for whom and when goes in the audit log. Only roles you allow can search, and staff can see searches about themselves.
Four things your cameras could tell you.
What a project can include, chosen with you for your site. The preview beside them shows how the dashboard is designed to look, with made-up numbers.
AI visitor analytics
Walk-ins counted at the door as they come and go, hour by hour, without anyone keeping a tally.
- Automatic visitor and walk-in counts
- Entry and exit traffic, door by door
- Daily, weekly and monthly visitor reports
- Peak hours picked out for you
Smart waiting-room monitoring
How full the waiting area is, roughly how long people wait, and a nudge to the desk before it overflows.
- Waiting-room occupancy estimates
- Approximate waiting times
- Queue congestion monitoring
- Capacity alerts at a level you set
AI operational intelligence
How busy reception and each area are through the day, so staffing can follow the traffic.
- Reception activity: visitors at the desk and the queue in front of it
- Occupancy by area: entrance, lobby, waiting room
- Traffic patterns across the day and the week
- Congestion alerts and insights for managers
Automated AI reports
A plain-language summary of each day and week, with the one or two things worth changing.
- Daily and weekly operational summaries
- Visitor traffic trends
- Peak-hour reports
- Recommendations you can act on, such as when to open a second counter
Visitor analytics
SimulationVisitors today
103
Waiting now
9 / 12
Approx. wait
21 min
Busiest hour
11 to 12
Entries so far
103
Exits so far
71
Busiest hour
11:00 to 12:00
Demo data. Tuesday, 12:30.
A simulated preview with made-up numbers, not a live dashboard.
How it connects to the cameras you have.
No new surveillance system, if your cameras can share a stream. The last two steps use parts of ResortOS that exist today.
Your cameras
The IP cameras and recorder you already have. We check each one before promising anything.
A read-only link
The live stream is read over RTSP, with cameras found through ONVIF. Camera and recording settings are left alone.
Analysed on site
A small computer at your premises runs the AI model. People are seen as moving shapes, never as faces.
Only results leave
Counts, times, alerts and the answer to a staff search go to ResortOS, encrypted. The video stays where it was recorded.
Dashboard and alerts
Occupancy, waits and congestion on the dashboard, and an alert to the desk when an area is nearly full.
Sign-in, roles and the audit log are in ResortOS today.
Reports
A summary each day and week: traffic, peak hours, longest waits and plain suggestions.
Daily and period reports for resorts are in ResortOS today.
Will it work with my cameras?
Often, but not always. RTSP and ONVIF are open standards that many IP cameras and recorders support. We would look at what you have before promising anything.
- IP cameras with an RTSP stream
- The likeliest to work. Many IP cameras can share a live stream over RTSP while they keep recording as usual.
- ONVIF-compatible cameras and recorders
- ONVIF is the open standard for finding cameras and their streams on a network, across brands.
- Network video recorders
- Many NVRs re-share each camera's stream, so analysis can read from the recorder and the cameras stay untouched.
- Older analogue CCTV
- Coaxial cameras on a DVR may need an encoder or a hybrid recorder first. We would tell you before you spend anything.
- Placement matters more than brand
- Counting works best with a clear, angled view of a door or waiting area in steady light. We would check each view on site.
Privacy first
Built to count, not to watch.
The design keeps video where it is recorded and sends only numbers. It is meant for rooms full of people who have not agreed to be identified, and it treats them that way.
In clinics and hospitals: the waiting room, never the patient.
Counts and approximate waits only. Cameras in consulting, treatment or changing areas would not be connected, and no count is ever linked to a patient or an appointment.
Before any site goes live, we would go through CCTV notices and retention with you, with India's Digital Personal Data Protection Act, 2023 in mind. ResortOS does not give legal advice.
Analysed where it is recorded
Video is processed on a device at your premises, not uploaded somewhere to be analysed.
Footage stays with you
Only counts, times, alerts and staff-search answers are sent to ResortOS. Recordings stay on your own recorder, under your own retention.
No faces
No facial recognition, for anyone. Visitors are counted as anonymous shapes; nobody is enrolled or matched by face.
Guests and patients stay anonymous
Guests and patients are never identified or followed. Staff can be located only on a manager's request, on shift, by the badge on their uniform.
Retention you control
You choose how long counts and alerts are kept, and they are deleted when that time runs out.
Access by role
The dashboard sits behind ResortOS sign-in, with roles, two-factor sign-in for owners and an audit log, as the product has today.
Where it fits.
Anywhere people arrive, wait and are served. Examples of how it could be used, not customers.
Healthcare
Clinics, dental and diagnostic centres, hospital outpatient areas.
- Waiting-room occupancy at a glance for the front desk
- Approximate waits, so people can be told honestly how long
- An alert before the waiting area is over capacity
- Peak-hour reports for planning appointments
Hospitality
Resorts, hotels and restaurants.
- The check-in and check-out rush at reception
- Breakfast and restaurant footfall, for staffing
- Walk-in traffic by day of the week
- Side by side with the rest of ResortOS on one dashboard
Commercial
Shops, showrooms, offices and service centres.
- Footfall and walk-ins by the hour
- Queues at counters and service desks
- Busy and quiet hours, for rosters
- Weekly trends for each branch
What it will not do.
No faces, no guest identities
It will not recognise faces or build a database of people, and guests and patients are never identified, followed or profiled.
Not a staff-watching tool
No movement history, no timing of breaks, no scoring. Finding a team member answers one question, when a manager asks it, and the person is told.
Not a security system
It does not replace your recording, alarms or security monitoring, and it is not meant for investigating incidents.
Estimates, honestly labelled
Counts and waits are estimates. Crowds, light and camera angles affect them, so accuracy is measured at your site during the pilot and shown to you.
Common questions.
Guide: counting visitors with the cameras you have- Can we use ResortOS CCTV intelligence today?
- It is not a ready-made product you can switch on. We build it for your site as a project, starting with a pilot at one location, after checking that your cameras and recorder can work with it.
- Do I need new cameras?
- It is built for the IP cameras and network video recorders you already have, reading their RTSP or ONVIF streams. Older analogue cameras may need an encoder first; we would check your cameras before you spend anything.
- Does it use facial recognition?
- No. Visitors are counted as anonymous shapes. Nobody is enrolled, matched or identified by face, and guests and patients stay anonymous.
- Is the video uploaded to the cloud?
- No. Video is analysed on a device at your premises. Only counts, times, alerts and staff-search answers are sent to ResortOS; recordings stay on your own recorder.
- What does it measure?
- Walk-ins and visitors by hour, how full a waiting room or lobby is, estimated waiting times, congestion alerts before an area overflows, and a plain daily report.
- How does finding a team member work?
- When a manager asks, it looks for that person's staff badge in work areas, only while they are on shift, and answers with where they were last seen. The person is told, every search is logged, and there is no movement history.
- Which businesses is it for?
- Clinics and hospital waiting rooms, hotel and resort lobbies and front desks, and shops, offices and other places where people wait to be served.
See what your cameras could tell you.
Tell us about your site, your cameras and your recorder. We will look honestly at whether they would work, and what a pilot at one location would involve.