Protective equipment
Someone in a work area without a helmet or vest: PPE detection on existing cameras.
SquintPRO
Already have CCTV? See what each part of the system does, and how AI analytics turns recorded hours into alerts, without replacing your cameras.
Explainer Published 6 min read
Teams we work with on SquintPRO






A CCTV system has four parts: cameras, a network that carries the video, a recorder or management software that stores it, and screens where people watch. Most systems record far more than anyone views. AI video analytics adds a fifth part that watches every stream and raises an alert. Interkey provides that layer; it does not sell or install cameras.
| Part | What it does | What to ask when buying |
|---|---|---|
| Cameras | Capture the picture of an area: fixed, dome or pan-tilt-zoom, indoor or outdoor | Is it an IP camera? Does it support RTSP or ONVIF? How does the picture look at night and in glare? |
| Network | Carries the video from the cameras to the recorder: cables and network switches, often with power for the camera on the same cable (PoE) | Is there enough capacity for every camera at the quality you record? Is the camera network kept apart from the office network? |
| Recorder or management software | Stores the video and plays it back by camera and time: a DVR for analogue cameras, an NVR for IP cameras, or video management software (VMS) across several recorders | How many days does the storage hold at your current resolution? Can it pass the streams on to another system? |
| Screens | Where people watch live and recorded video: a control room or a security desk | How many cameras can one operator follow? Who may view and copy recordings? |
Most systems in use today have all four parts. What they lack is a part that watches.
Camera
captures the picture
Network
carries the video
Recorder or management software
stores the video and plays it back by time
Screens
where people watch
AI analytics layer
reads the streams from the cameras or the recorder, watches each one, and raises an alert
The first four parts record and show. The fifth reads the same streams and changes nothing in the other four.
An analogue camera sends its picture as a signal over a coaxial cable to a digital video recorder (DVR). An IP camera sends its video digitally over the network, where any authorised device can reach its stream.
That difference decides whether analytics is possible. AI video analytics reads streams over the network, so it works on IP cameras that support RTSP or ONVIF, the open protocols most cameras and recorders use. An analogue camera can join through an encoder that turns its signal into an IP stream, if its picture stays clear enough. Either way, a survey of each camera sets its fitness, and camera compatibility for SquintPRO lists what it checks. To weigh new AI cameras against software on the cameras you have, read AI cameras.
Before adding anything to a surveillance system, a site settles four questions in writing:
The requirements that apply to your facility are settled with your legal advisers; this guide does not interpret them.
The third part is the system's memory. On a small site one recorder is enough: a DVR for analogue cameras or an NVR for IP cameras. When sites and recorders multiply, video management software (VMS) gathers them in one place: live view, search of recordings by camera and time, and user permissions.
However large they grow, the recorder and the VMS keep what the camera sees without understanding it. Their search runs by time, so anyone looking for an event without knowing when it happened watches the hours themselves. Our guide to the video management system shows where its role ends and video analytics begins.
Relatedvideo management system
A system of 100 cameras running all day records 2,400 hours of video every day. One operator can follow only a few screens at once, and attention fades as a shift goes on. So recordings are usually opened after an incident, to learn what happened, not to prevent it.
The analytics layer changes that order. It watches every connected stream and sends an alert with a short clip when a defined event occurs, such as a missing helmet or smoke. The operator turns to the camera that deserves a look, instead of moving from screen to screen.
Someone in a work area without a helmet or vest: PPE detection on existing cameras.
A person too close to a crane, forklift or other moving machine: forklift and pedestrian detection.
Smoke or flame in the picture raises an alert with the clip: smoke and fire detection by camera.
A person or vehicle entering a zone in the hours you set: restricted area monitoring.
Plates read at gates and car parks, and a vehicle followed across cameras: number plate recognition.
Type what you are looking for instead of watching the hours: searching recordings by description.
Interkey's scope
SquintPRO works on the IP cameras you already have. Interkey does not supply or install cameras.
Interkey provides the analytics layer: the camera survey, connecting the streams, running on-premise or in the cloud, and tuning the alerts and delivering them to your control room.
A CCTV system has four parts: cameras, a network that carries the video, a recorder or management software that stores it, and screens where people watch. It is used to protect a site and to review what happened after an incident. Most of what it records is never viewed, which is where AI video analytics comes in: it watches the streams and raises an alert.
Four core components: cameras that capture the picture; the network of cables and switches that carries it; the recorder (a DVR or an NVR) or video management software that stores it and plays it back by time; and the screens in a control room or at a security desk. AI video analytics adds a fifth component that reads the same streams without changing the other four.
IP cameras send video digitally over the network, and they are the cameras analytics works on; analogue cameras send a signal over a coaxial cable to a DVR. An analogue camera can still be connected to analytics through an encoder that turns its signal into an IP stream, once the survey confirms that its picture stays clear enough.
Yes, if the cameras are IP cameras that support RTSP or ONVIF, and a survey confirms which ones qualify. SquintPRO reads the streams from the cameras or from the recorder, recording stays on your system as it is, and alerts reach your control room. The survey also shows which cameras need a change of position or lighting.
No. SquintPRO works on the IP cameras you already have, and Interkey does not supply or install cameras. Interkey provides the analytics layer on top of your system: the camera survey, connecting the streams, running on-premise or in the cloud, and delivering alerts to your control room. If a camera needs replacing, the survey says so in writing for your installer.
Next
Next step
Tell us how many cameras you have, what type they are and what records them, then ask one question about adding analytics.
What happens next
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