Watch the boundary and the empty buildings, not the people.

A school or campus is busy during the day and empty for most of the week. The security problem is concentrated in the hours nobody is there, and at the boundary — which is also where a small team has the least ability to watch.

What makes this estate hard to watch

Empty far more than it is occupied

Evenings, weekends and holidays are the exposure. Schedules matter more here than almost anywhere else.

A large boundary and a small team

Sites are open in shape and often lightly staffed outside hours, with one caretaker covering a lot of ground.

Mixed, inherited camera estates

Cameras added in phases over many years, with uneven coverage and uneven night performance.

From the trigger to the evidence

Which of these apply to you, and which cameras can carry them, is what the site survey establishes before anything is proposed.

01

Someone on site at night

Trigger

A person detected in a marked outdoor area during hours the site is closed.

Camera context

The detection frame with the area drawn on the camera view.

Operator review

A person or an animal, settled from the picture rather than a sensor state.

What happens next

Light on where one is wired, voice challenge where a speaker covers it, alert to the caretaker.

02

Entry to a plant room or store

Trigger

A person inside a restricted area outside its permitted window.

Camera context

The frame at the moment of detection.

Operator review

Contractor working late, or an unexplained entry.

What happens next

Alert with the clip, retained for the incident log.

03

Activity at a boundary or gate out of hours

Trigger

Movement at the fence line or a gate when the site is closed.

Camera context

The boundary camera, lit if a floodlight is wired in.

Operator review

Whether it warrants a call-out.

What happens next

Escalation to the nominated phones with the frame attached.

04

A camera covering an exposed area stops working

Trigger

A camera goes offline, is moved or is obscured.

Camera context

The last usable frame and the time it stopped.

Operator review

Raised as an event in its own right.

What happens next

Repaired before an incident proves it was blind.

Built around what you already run

Configured around hours, not people

Areas carry the schedule they are off limits. During the day, those areas simply do not raise events.

Uses the cameras already there

Standard RTSP and ONVIF, read directly or through the existing recorder.

Local processing

The appliance runs on your own network. Continuous video does not leave the site.

Night performance is a hardware question

An unlit field at the end of a wide lens cannot be fixed in software. The survey says which areas need light before anything is promised.

The configuration here is deliberately about places and hours rather than individuals. What is retained, for how long and who can see it are decisions you make and we configure, and they belong in your own policy review before anything runs.

One appliance. NVR + AI + management.

Camera streams, recording, playback, AI events, evidence clips, alerts, reports, users and roles and camera health — centralised in one appliance inside your premises. Faces, clips and reports stay on site.

Input IP / Wi-Fi / compatible existing cameras
Network PoE / network
On premises PARVEK Appliance Recording · AI detection & rules · management
Mobile alertVoice warningLight / floodlightSiren / hooterEvidence clipReports
The PARVEK ops console: an event list on the left, the selected UNKNOWN PERSON detection frame in the centre, and resolve, escalate and physical-response controls on the right
Ops console — events, evidence and response in one place
The PARVEK mobile app showing a live camera view, an SOS/panic button, and tiles for devices, notifications, alerts and recordings
Mobile app — live view, alerts and SOS
Camera health diagnostics listing eight cameras with their frame rates, all green
Camera health — offline and degraded cameras flagged

Camera count, retention and storage are specified per appliance configuration and confirmed during the assessment.

What to measure on your own site

We do not publish improvement percentages, because none have been measured across a fleet and yours would not match them anyway. What a bounded proof of concept gives you is your own numbers, on your own footage:

  • Events per night across the boundary and outdoor areas
  • How many need a call-out versus how many are wildlife or weather
  • Which cameras give a usable image after dark, and which need lighting
  • Time from detection to the caretaker or on-call person having the frame
  • Whether the response — light, voice, call-out — is right for each area

Frequently Asked Questions

Are you monitoring students or staff?

No. The useful configuration for a campus is areas and hours — a boundary at night, a plant room out of hours. During the school day those areas raise nothing. We would steer you away from anything that amounts to monitoring people going about their day.

Do we need to replace our cameras?

Usually not. What matters is placement, lighting and resolution at the distance involved rather than age. The free camera check and site assessment tell you which of yours are usable before anyone proposes a purchase.

Can it detect weapons?

Threat detection is offered only on qualified deployments after site-specific validation, and it is not something we enable because a tier was purchased. For most campuses the questions that actually get answered are boundary and out-of-hours presence.

Who gets the alerts at 2am?

Whoever you nominate — caretaker, on-call manager, or several people at once. Delivery is push notification with the frame attached, plus the console for anyone at a desk.

AI surveillance and analytics are technology-assisted and do not guarantee prevention or detection of every security incident or 100% accuracy.

Start with a site survey

Tell us what the site is, what it runs and what you need it to catch. The survey establishes whether your cameras can carry the use case — and from there, a proof of concept runs it on your own footage.