Cloud VMS for security agencies: a service on the cameras your clients already own
By Muhammad Talha · Product Manager and CTONine years building computer vision and automated surveillance systems · Updated 
A cloud VMS lets a guarding company, a monitoring center or a CCTV installer sell a monitored video service on cameras the client already has, from one account, with a login per client. The cameras stream to the cloud instead of to a recorder on site, and the rounds, alerts, recordings and reports for every client are run from one console. This guide answers the questions an owner asks before putting that service on a proposal: how the accounts separate, how per-camera billing maps to a monthly price, what the client sees, what you need on your side, and how to start with one client.
Can a cloud VMS turn my clients' existing cameras into a monitored service?
Yes, provided the cameras produce a stream the service can read, which most IP cameras from the last decade do. Cloud video surveillance takes an RTSP, RTMP or HTTPS stream from each camera and records, watches and reports on it in the cloud, so the client's recorder can stay exactly where it is.
Camzify is software only. There is no hardware to stock, ship or support, and no preference for any camera manufacturer. A camera reachable from the internet is added by its stream address. A camera on the client's private network is relayed by the Camzify Connector, a small application on a Windows, macOS or Linux machine inside that network, so there is no port forwarding to negotiate with the client's IT.
What you are selling on top of that stream is the part a recorder cannot do: scheduled patrol rounds with a checklist at each camera, AI detections between rounds, a guard notified when something fails, and a report per round the client can hold. The virtual guard page describes that service in the market's own words. Camzify is the software; your company is the guard the client contracts with.
How do accounts and permissions work when I have many clients?
Your company holds one account, and each client is a sub-user scoped to their own sites and cameras. Sites are the unit everything is organized around, so access, reporting and alert filtering follow the site, and a client logging in reaches nothing that belongs to another client of yours.
A permission group decides what each login can open and change. It combines page-level access with create, read, update and delete rights per resource, and one group is assigned per user. The ready-made Auditor group sees everything and changes nothing, which suits a client checking on the service they pay for. User management covers both halves in detail.
- License quota is allocated to each client from your own: AI feature instances, patrolling instances and backup storage, never more than you hold.
- When a client reaches a limit, their request comes back to you to approve or decline rather than being granted automatically.
- Sub-users can create their own sub-users, so a client can give its own staff a login without involving you.
- A sub-user can be deactivated without being deleted, keeps its sites, recordings and reports, and can be reinstated.
The same model works when the parties are stacked differently. A monitoring company running rounds for several agencies holds the account and gives each agency a scoped login; the monitoring companies page covers that three-party arrangement, including who receives the guard message.
How does per-camera billing map to what I charge?
Camzify is priced per instance per month and quoted for the portfolio you would cover; you then set your own price per client site per month against that quote. The two numbers are yours to keep apart, and your proposal should not repeat ours.
The quote is built from what is switched on where. The platform modules come with the account. Every connected camera takes a stream instance, and each AI detection feature has a pool of instances, activating it on a camera consumes one, and patrolling instances are counted per camera per sequence. A corridor camera can record and nothing more while the gate camera records, detects and is a stop on the round, so a client site does not have to carry every feature on every camera.
Camzify starts from $5 per camera per month and a partner portfolio is quoted as a whole; there is no published partner margin, tier or portal. Terms are agreed in conversation, and the reseller page says so rather than inventing a program. The ROI calculator has an agency mode that takes your client sites, the price you would charge and the hours you cannot staff, and shows the recurring revenue; the margin is the difference between that and your quote.
What does the client see, and what do I see?
The client sees their own sites through their own login: their live cameras, their alerts and their patrol reports, filtered by the permission group you gave them. You see every client on the parent account, on one live wall grouped by site with a per-site online count, with every alert in one queue and every report under the site it belongs to.
Whether a client gets live view or the report only is a commercial choice, and both are supported. Some agencies give the client the Auditor view so they can read reports and watch cameras without changing anything. Others hand over reports and keep the console to themselves. The client's login works in the browser on a phone as well as a desktop, with nothing to install.
Alerts are delivered by email, SMS, WhatsApp and push, with channels configured per alert category and severity set per camera for each detection. Each detection has a notification window per camera, so a person in a loading bay notifies at 02:00 and stays quiet at 14:00 while the detection itself runs all day.
How do patrol rounds and reports become a deliverable?
A patrol round is an ordered list of camera stops with a checklist at each, run on a schedule, with a report filed at the end. The report is the deliverable: it is what the client receives after every round, and it is the part they cannot get from anyone who does not run rounds.
Automated rounds run on a schedule set by frequency, active hours and active days in the site's own timezone, and each client site can have its own sequence and schedule. At each stop the round marks every checklist item from the frame. An item found Not Compliant captures a snapshot and messages the guard assigned to that camera, with no operator approving the message. A manual round offers the message to the operator instead.
- The report names the site and lists the camera stops in order.
- It records every checklist result and the snapshot each result was judged against.
- It carries the before-and-after pair for anything that failed and was fixed during the round.
- It carries timestamps, the guard notified on any failure and an overall compliance percentage.
A round cannot be closed with an item still Not Compliant, and a Pending item counts against the percentage, so the number means what it says. The report opens as a web report or a PDF, and the client reads it through their own login or you send it. Nothing in it is typed in after the fact.
What does my agency need on its side?
Three things: a person who owns the console, a written workflow for what happens when a check fails, and enough upstream bandwidth at each client site to carry the cameras you stream. None of them is hardware, and none of them is a control room if you do not already run one.
The console owner sets up sites, builds the sequences and checklists, assigns the guard to notify per camera and creates the client logins. Automated rounds then run with nobody in the loop, so the day-to-day load is the notifications, not the rounds. For an agency that sells mobile patrols today, that is usually the dispatcher or the operations manager.
The workflow is the part most owners skip. Decide before the first round who receives the guard message for each client site, what they do with it, and who closes the item as Fixed so the report carries the second snapshot. A staffed monitoring center adds manual rounds for the checks it wants a person to make and takes the critical notifications the automated round raises.
Bandwidth is the honest constraint. Cloud video surveillance needs upstream bandwidth for every camera streamed, and a site that cannot carry its cameras, or whose policy says footage never leaves the building, should keep its recorder. Check the uplink before you promise the service, and stream the cameras that are stops on the round rather than every camera on the site. A CCTV installer you already work with can answer that question in a visit.
How do I start with one client?
Pick the client whose cameras you can reach and whose overnight gap you already know about, and run one round on their cameras before you write a proposal. A report with the client's own dock door in it settles the question of whether the service is real, and it gives you the sample to price against.
- Book a demo and we will run a live round on cameras you can reach, with you and again with the client if that helps.
- Create the client as a sub-user scoped to their site, choose a permission group, and allocate the instances the round needs from your license.
- Build one sequence with a checklist per camera stop, assign the guard to notify, and set the schedule for the hours you cannot staff.
- Price it per site per month at your rate, beside the sample report and the schedule, and keep the guard hours in the contract as response.
The security agencies page covers the commercial argument and the account model in full, and the selling guide below covers the client conversation. What we will not tell you is that this replaces your guards, or what your response time or detection rate will be; the trust page lists what we do not claim, and a proposal that stays inside those lines is one the client will not catch out later.