The trigger moments that start this conversation, what actually changes on day one, and a four-phase plan that keeps every site recording throughout. Written for teams running an on-premises video management system on their own servers who want more from the cameras already on the wall.
The cameras do not move, so switching changes what sits behind them and what the recording is asked to do. The before row describes the common on-premises pattern.
Four moments when teams running an on-premises video management system start this conversation. If none of them are true, staying put is a reasonable answer.
Recording servers reach end of support and the estate needs a capital decision anyway. That is the cheapest moment to ask whether the replacement should be the same architecture with newer hardware, because the budget is approved and the disruption is already scheduled. Model the next three years both ways against the same camera count and the same retention policy.
Per-channel licensing plus an annual maintenance agreement renews on a date somebody in finance already has in a spreadsheet. Renewal is the honest moment to compare, because the platform decision is open regardless. Ask for both quotes broken out into software, recording hardware, storage, installation and support.
A traditional VMS answers what happened once somebody knows roughly where to look and has time to scrub. When the security team can name the last three investigations that cost half a day each, the business case stops being about the recorder and starts being about the search. Liberty-Perry School District reports resolving an incident in about five minutes.
One building with one recording server is manageable. Twenty buildings means twenty upgrade windows, twenty storage calculations and a user list per site. If the answer to who has access to site fourteen takes an email to find, the administration model is the constraint rather than the software.
The structural differences that decide whether a switch is worth starting.
Swipe the table sideways to see every column.
Traditional VMS is a category rather than a company, so this column describes traits common to on-premises video management systems rather than any single vendor's documentation. Check every line against your own product's release notes and license terms before you plan around it.
A rollout that keeps coverage live throughout. Most sites go live inside six weeks.
Inventory the fleet per site: manufacturer, model, resolution, codec and network path, plus which channels are analog sitting on encoders or a hybrid recorder. Then put three dates next to it, because they decide the sequence: when each recording server leaves support, when the maintenance agreement renews, and how long the retention policy actually requires. Because Spot AI reads ONVIF and RTSP, this phase is a coverage map rather than a replacement list.
Run one site live on both systems at once. The existing servers keep recording while the same cameras stream to an Intelligent Video Recorder, which is the only way to compare search, detection quality and what happens after a detection on identical footage. Pick the site with the messiest camera mix rather than the newest, because a fleet assembled over ten years is the real test and the newest site flatters everybody. Watch the noise as well as the detections: a leading SOC platform integrated with more than 20 VMS platforms reports a single-digit false-positive rate against about 50 percent false alarms on legacy systems, and a fortnight on your own site is how you check that against your own numbers.
Move site by site rather than estate-wide, keeping the existing recording live at each site until its replacement has run a full week. Cameras do not move and cabling does not change, so a site cutover is a configuration change plus an IVR, not a re-survey. Retire a recording server only once the site behind it has been signed off, and let the sites whose hardware is oldest go first.
Once coverage is stable, the work moves from parity to what a recorder was never asked to do: adding operations and safety agents on the same cameras, building an Iris detection for something specific to your process, and setting what fires at 03:00 when nobody is watching. This is the phase that turns a hardware refresh into an operations project, and it is where the switch stops being a cost exercise.
The six that come up in every one of these conversations.
Perpetual licenses are a sunk cost, and the money spent on them does not come back whichever way this goes. The number that matters is what the next three years cost from here: maintenance renewals, the server refresh already on the plan, storage growth and the hours investigations take. Compare that against the alternative rather than against what has already been spent.
No, and the plan is built to make that structural rather than a promise. The existing servers keep recording at each site until the replacement has run a full week, the cameras never move, and cutover happens site by site. If a site fails its sign-off test, that site stays where it is while the others proceed.
It stays on the storage it is on now, under the retention policy you already run, and it does not migrate. Export anything under a legal or insurance hold before the servers are retired, in the format your evidence process expects. Treat the export list as a deliverable of the assess phase, because it is the one task that gets forgotten until the hardware is gone.
That policy survives the switch. Full-resolution video records to the Intelligent Video Recorder on site and only event metadata leaves the building, which is the same boundary an on-premises VMS gives you today. Put the policy in front of both vendors in writing during assess, because it is cheaper to have that conversation before a pilot than during a security review.
It depends on your retention, your site count and how much of the estate is due for refresh anyway, and anybody who answers without those numbers is guessing. What changes is the shape of the spend: capital every few years for servers and storage against a per-camera subscription covering the software and the IVR. Quote both against the identical site list and the identical retention obligation over three years.
Six things, and they are the same six whichever way you go. When each maintenance agreement ends and what notice it needs. Whether licenses are perpetual, co-terminated or per device. The export format and any time limit on historical footage. Which cameras are still under warranty. Who owns the camera VLAN at each site. And your own retention and data-residency policy in writing, not from memory.
Bring a camera list and we will tell you what works as it is.
Customer-reported outcomes from named Spot AI customers.
Staccato went from first conversation to full deployment across an 800-acre campus in seven weeks, on the camera infrastructure it already had.
Bridge33 Capital standardized video across 25 plus properties, each acquisition arriving with a different camera system, and cut footage search from hours to minutes.
The YMCA of Greater Richmond deployed 17 locations in two weeks and standardized retention across the estate.
"Within four minutes of the alarm going off, Spot AI gave us video footage of the incident on the responding officers' phones. We had five recoveries within the hour."
No. Spot AI reads any IP camera that speaks ONVIF or RTSP, which is the same standard your VMS uses to talk to them today. Analog channels currently running through encoders or a hybrid recorder connect through the Intelligent Video Recorder instead. Nothing on the wall moves and no coax is pulled.
Most estates are live inside six weeks: a week to assess, a fortnight of pilot on one site, then site-by-site cutover. Sites go live individually rather than in a big-bang switch, so the timeline scales with how many buildings you have rather than how many cameras. Staccato reports seven weeks from first conversation to a full 800-acre campus.
It stays where it is under your current retention policy and does not transfer. Anything under legal or insurance hold should be exported before those servers are decommissioned, in the format your evidence process expects. That export list belongs in the assess phase rather than in the week you are unracking hardware.
Yes, and the plan depends on it. A camera can stream to the existing recording server and to an Intelligent Video Recorder at the same time, so the pilot compares search, detection and response on identical footage. The old system keeps recording at every site until that site has passed a full week of sign-off.
A traditional VMS is a recorder with a client on top: it captures reliably and waits to be asked. Spot AI keeps the same on-site recording model, with full-resolution video on the Intelligent Video Recorder and only metadata leaving the building, and puts pre-trained agents and plain-English search on the footage so the system raises what matters instead of waiting for somebody to go looking.