Try it on your own cameras, with the appliance on loan
Thirty days on your own site, your own cameras and your own footage. We lend the appliance at no cost, commission it with your team, and collect it if you decide not to continue.
PARVEK runs on an appliance on your premises, so a real trial needs one there. We lend it for the 30 days. You are evaluating detection on your own site — not buying a box to find out whether the software is any good.
What you get for the 30 days
The appliance, lent at no cost
We provide the PARVEK appliance for the trial. You do not buy hardware to evaluate the software. If you decide not to continue, we arrange collection and that is the end of it.
Your cameras, not ours
The trial runs on the cameras already on your wall, so what you are judging is detection on your own scenes, lighting and angles — not a vendor demo on footage chosen to flatter the model.
Installation and configuration with you
We commission it alongside your team: camera connections, the zones that matter, the hours they matter, and who gets alerted. This is the work that decides whether a deployment is any good.
The capabilities your site needs
Which ones run is decided at the assessment rather than fixed by a plan. Capabilities that need more validation, such as face recognition or specialized models, are agreed before the trial starts.
Your access-control system connected
If you run a PACS, we build the event connection during onboarding and correlate door events against your cameras inside the trial — so you watch access verification work on your own doors before committing to it.
Tuning on your own footage
No site is right on day one. We review what fired and what did not during the trial and adjust thresholds, zones and schedules against your real events.
A written read at the end
What worked, what did not, which cameras are holding you back, and what a full deployment would involve. You keep that whether or not you proceed.
What makes a trial worth running
A trial is only useful if it tests the real thing. These are the conditions that make it a fair test rather than a demo.
We need to know which of your cameras can carry detection before anything is shipped. That is the two-business-day assessment, and it is free.
The appliance needs power, a network path to the cameras, and somewhere to sit. Outbound connections only — no inbound ports, no public IP.
A person who knows the site and can say which areas matter and when. A trial with nobody to tell us what counts as an incident will not tell you much.
If your cameras cannot see what you want detected, a trial will show that rather than fix it. We would rather tell you at the assessment than after 30 days.
Two outcomes, both of them fine
The appliance stays and moves onto the normal per-camera subscription at the tier your site actually needs. Nothing is billed for the trial period.
See what it would cost →We arrange collection. No purchase, no deposit, no automatic rollover onto a paid plan, and the data on the appliance is destroyed. You keep the written assessment.
Frequently Asked Questions
Do we have to buy the appliance to try it?
No. The appliance is lent at no cost for the 30 days. If you do not continue, we arrange collection — there is no purchase, deposit or restocking charge for the trial itself.
What happens at the end of the 30 days?
One of two things. You continue, and the appliance stays and moves onto the normal per-camera subscription; or you do not, and we collect it. There is no automatic rollover onto a paid plan, and nothing starts billing unless you say so.
Does our data leave the site during the trial?
No continuous video leaves your premises during the trial, exactly as in a full deployment. Analytics run locally on the appliance. Alerts and short clips reach the people you nominate, over outbound connections only.
What happens to the data if we do not continue?
It is on your appliance, on your premises. When the appliance is collected the data on it is destroyed, and we can confirm that in writing if you need it for an internal sign-off.
Can we trial on part of the site?
Yes, and on a large estate it is usually the sensible route — run detection where incidents actually happen, judge it there, then extend. The assessment establishes which area is the useful test.
Can we try the access-control integration in the trial?
Yes — that is the point of doing it in a trial. We confirm what your access-control system can export, build the connection during onboarding, and you see door events correlated against your own cameras. Identity matching additionally needs face recognition enrolled and your PACS users mapped to enrolled people, which we scope before the trial starts.
Do we need new cameras?
Usually not. The assessment confirms which existing cameras qualify and says plainly which do not. Where a camera genuinely cannot see what you need, we say so before the trial rather than letting it fail quietly during it.
How soon can it start?
The assessment comes back within two business days. Scheduling the install depends on your site and ours — tell us your timeline at contact@parvek.ai and we will be straight about what is achievable.
AI surveillance and analytics are technology-assisted and do not guarantee prevention or detection of every security incident or 100% accuracy.