SquintPRO · Camera compatibility
AI video analytics for existing CCTV: SquintPRO on the cameras you already own
Most organisations in Saudi Arabia already have hundreds of cameras recording around the clock, and almost nobody watching them. AI video analytics for existing CCTV adds the watching without replacing the estate. SquintPRO, the video analytics platform Interkey builds, connects to the IP cameras and recorders you run today over RTSP and ONVIF, analyses the streams on-premise or in the cloud, and sends alerts with video evidence into the system your team already uses. This page explains which cameras qualify, what the site needs for each deployment model, and how a project runs.
In short
- What it is: SquintPRO, Interkey's video analytics platform, on the CCTV you already have.
- How it connects: to IP cameras and network video recorders (NVRs) over RTSP and ONVIF; older analogue cameras connect through an encoder.
- What decides compatibility: a camera survey of each view, not a model list.
- Where it runs: on-premise, so footage stays on your site, or in the cloud with 3, 6 or 12 months' retention.
- How it is bought: by quote, licensed per camera stream, with unlimited use cases on a licensed stream.
- What it does not do: identify people. There is no face recognition.
At a glance: what SquintPRO needs from your camera estate
| Item | What SquintPRO needs | What the survey checks |
|---|---|---|
| IP cameras | A video stream it can reach over RTSP, or a camera it can discover and configure over ONVIF | That each stream opens, and which of its profiles gives enough detail for the use case |
| Analogue cameras | An encoder that turns the analogue signal into an IP stream | Whether the picture is clear enough once encoded |
| NVR or VMS | Access to the streams the recorder already receives, where it can pass them on | Whether the recorder can re-stream without overloading it, or whether streams are taken from the cameras |
| Network | A path from the cameras or recorder to the processing hardware | Bandwidth and segmentation on that path |
| On-premise processing | A GPU server on site, new or on existing infrastructure, with network access to the cameras | How many streams and use cases it must carry |
| Cloud processing | A unit on site that collects the streams, and a link able to carry them out of the site | Whether the site link can carry the streams in scope |
| Alerts | A destination: the VMS, control room or incident system already in use | Which integration the destination supports |
Each alert carries a short clip so an operator can verify it.
What "works with existing CCTV" has to mean
For a buyer, "works with existing cameras" should mean three things, each checkable before a contract:
- The connection is standard. SquintPRO reads streams over RTSP and finds and configures cameras over ONVIF, the open protocols most IP cameras and recorders support. No proprietary camera is required.
- The cameras are kept. Compatible cameras are not replaced. Where a view cannot support a use case, the survey says so and proposes repositioning, lighting or, only where needed, a new camera.
- The workflow is kept. Alerts arrive in the VMS and incident process your operators already watch, not on another screen they must learn.
Read more
What the phrase cannot promise is that every camera suits every use case. A wide overview camera on a car park can show a vehicle entering a restricted lane, but rarely gives a readable number plate. Compatibility is a question about each view and each use case, which is why the first step is a survey.
How SquintPRO connects to IP cameras, NVRs and the VMS
There are two ways to bring a stream to SquintPRO, and most sites use a mix of both.
- From the camera. SquintPRO opens the camera's RTSP stream directly. Where a camera offers a high-resolution main stream and a lighter sub-stream, the survey picks the one with enough detail for the use case.
- From the recorder. Where the NVR or VMS already receives every camera and can pass streams on, SquintPRO takes them from there, so the cameras are not asked for an extra connection. The survey checks that the recorder has the capacity to do this.
Read more
Recording stays where it is. SquintPRO does not replace your NVR or VMS; it adds analysis and sends each alert back to them or to your control room. Connected streams can also be searched with AI video search, by describing what you are looking for or by supplying an image.
What the site needs: on-premise or cloud
| Point | On-premise | Cloud |
|---|---|---|
| Where the video is analysed | On a GPU server at your site | In the cloud, after a unit on site collects the streams |
| Does footage leave your site? | No | Yes, the streams in scope are sent for analysis |
| What you provide | Rack space, power and network access from the server to the cameras or recorder | A link from the site able to carry the streams in scope |
| Retention of analysed footage | Set by your own storage | 3, 6 or 12 months, chosen at the start |
| Usually chosen when | Policy says video must not leave the site, or the link cannot carry the streams | You prefer not to run processing hardware on site |
Interkey sizes the server to the streams and use cases in scope after the survey. How the choice moves the quote is explained on what determines SquintPRO pricing.
What AI video analytics for existing CCTV adds
Once a stream is licensed, use cases on it are unlimited, so one camera can carry several at once. Typical starting points on existing cameras:
- Restricted areas and perimeters: an alert when a person or vehicle enters a zone during the hours you set; see intrusion detection on existing cameras.
- Vehicles on site: vehicles where they should not be, followed across cameras; see vehicle detection on existing CCTV.
- Industrial safety: missing protective equipment, people too close to moving equipment, entry into hazardous zones.
Start with the one or two use cases that matter most, proven on your own footage first.
Relatedintrusion detection on existing camerasvehicle detection on existing CCTV
What Interkey delivers
Interkey builds SquintPRO and carries out the project around it: the camera survey, the stream choice for each camera, the server or cloud setup, the integration with your VMS and alerting, tuning, and operation after go-live. Support comes from Riyadh, with engineers who work in Arabic and English alongside the people who run your control room.
For organisations in the Kingdom that hold video under a data policy, the deployment choice often comes first. On-premise keeps footage on your own site, inside Saudi Arabia. If your sector regulator or internal policy sets rules on where video may be held, name them at the start so the design follows them; for national cybersecurity requirements, see the National Cybersecurity Authority.
How a project runs
-
Camera survey.
Each camera is sorted, per use case, into usable as is, usable after adjustment, or needing new hardware.
-
Pilot scope.
One site, a set of cameras, one or two use cases, and written success criteria, including the false alerts per shift you accept.
-
Connection and processing.
Streams are connected over RTSP or ONVIF, from the cameras or the recorder, and the on-premise server or the cloud setup is put in place.
-
Integration and tuning.
Alerts reach your VMS or control room, and each use case is tuned on your footage.
-
Measurement and rollout.
Results are measured against the written criteria. More cameras, use cases or sites are added once the pilot passes.
What moves the cost of adding AI to existing CCTV
SquintPRO is licensed per camera stream, with unlimited use cases on a licensed stream, and each quote is built from the site, the number of camera streams and the deployment model. Four things move the total:
- Streams in scope. Only connected streams count; start with your highest-risk areas.
- On-premise or cloud. On-premise needs a GPU server on site; cloud needs a link that can carry the streams and a retention period.
- Camera work. Repositioning, lighting, encoders, or an occasional new camera.
- Integration depth. One VMS is lighter than several systems and sites.
The SquintPRO questions page answers common questions on privacy, placement and accuracy.
Questions buyers ask
Which of our current CCTV cameras can SquintPRO analyse?
Most IP cameras that provide an RTSP stream or support ONVIF can be connected, and older analogue cameras can be connected through an encoder. Whether a camera suits a use case depends on its angle, the size of the subject in the image and night lighting; the survey lists, in writing, which cameras are usable as is and which need adjustment.
Can SquintPRO take the video from our NVR instead of connecting to each camera?
Yes, where the NVR or VMS can pass on the streams it already receives. The survey checks the capacity of the recorder; where it is short, streams are taken from the cameras directly. Recording stays on your existing system.
What hardware do we need on site for on-premise video analytics in Saudi Arabia?
A GPU server at your site, new or on existing infrastructure, with network access to the cameras or the recorder, plus rack space and power. Interkey sizes it after the survey. Footage is analysed there and does not leave your site.
If we choose cloud processing, does our CCTV footage leave the site?
Yes. A unit on site collects the streams in scope and sends them for analysis, and recordings are kept for 3, 6 or 12 months, chosen at the start. If your policy or regulator requires video to stay on your premises in the Kingdom, choose on-premise instead.
How is adding AI to existing CCTV licensed and priced?
By quote. SquintPRO is licensed per camera stream, with unlimited use cases on a licensed stream, so adding a use case on a licensed camera does not add a licence. See what determines SquintPRO pricing.
How long does it take to put analytics on our existing cameras?
It depends on the number of cameras, how quickly site and network access are arranged, whether on-premise hardware must be bought, and how many systems the alerts must reach. Interkey gives an estimate after the survey; the rollout is planned once the pilot meets its written criteria.
Next
Related pages
Next step
Check our cameras for AI analytics
Tell us roughly how many cameras you have, what records them, and which risks you want to watch. We will tell you which cameras are likely to qualify, what the site would need for on-premise or cloud, and what a pilot could measure.
Or directly
+966-11-2180999 [email protected]Tawuniya Towers, North Tower, 7th Floor, King Fahad Highway, Olaya, P.O. Box 56835, Riyadh 11564, Saudi Arabia