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 security operations teams running an intelligence layer over the systems they already own.
One platform layers over the systems you already run and the other is the video platform itself. Neither asks you to replace a camera, so the difference is where the frame gets understood.
Four moments when teams running a monitoring intelligence layer start this conversation. If none of them are true, staying put is a reasonable answer.
Cobalt reasons over access-control events, video triggers and sensor alerts, and it publishes real numbers for that job: at Salesforce it reports returning over 6,000 hours of operator time a month. The trigger is a different question arriving. When EHS or operations start asking what the cameras themselves can tell you about PPE, dock dwell or an SOP step, that is a video-platform decision rather than an alarm-triage one.
Cobalt layers over your existing infrastructure with no overhaul required, which means it sits on top of a video system of record you are also paying for. When both renewals land in the same year, price the combined stack against one platform that records and reasons in the same place, using the identical site list. Sometimes the stack wins. Do the arithmetic rather than assuming either answer.
Cobalt publishes its automated actions clearly: service tickets, access-control updates, Slack, SMS and guard dispatch, with operators making the call on the events that need one. Speaker, talk-down, strobe or horn behavior is not publicly specified. If your overnight problem is people in a yard rather than a door sensor reporting a fault, ask both vendors to describe the sequence at 03:00 before anybody arrives.
An intelligence layer inherits whatever the underlying platform gives it: its retention, its search, its camera compatibility, its bandwidth profile. If the constraint you keep hitting is in that layer rather than in the reasoning above it, adding more intelligence on top will not move it. That is the case where replacing the video platform is the cheaper project.
The structural differences that decide whether a switch is worth starting.
Swipe the table sideways to see every column.
Cobalt AI details come from Cobalt's own site and platform pages, read on August 14, 2026. Where Cobalt publishes nothing, this table says so rather than guessing.
A rollout that keeps coverage live throughout. Most sites go live inside six weeks.
Two inventories, not one. First the cameras per site: manufacturer, model, resolution, codec and network path, which is a coverage map rather than a replacement list because any ONVIF or RTSP stream connects to the Intelligent Video Recorder as it is. Second the event sources Cobalt currently reasons over, split into the ones that come from cameras and the ones that come from access control, sensors or device health. The second list is what decides whether this is a replacement or a narrowing of scope, and it is better drawn before anybody quotes.
Run one site live on both at once. Cobalt keeps reading the systems it reads today while the IVR records the same cameras, so nothing is switched off to run the comparison. Pick the site whose overnight events are about people and vehicles rather than door faults, because that is where a platform that reasons about the frame and one that reasons about the alarm behave differently, and because it is the comparison your operators can actually judge.
Move site by site rather than estate-wide, and treat the two layers separately. The cameras and the recording move first, per site, with the previous video platform live until its replacement has run a full week. What Cobalt does with access control, device health and the alarm queue is a second decision with its own date, and there is no requirement to make both in the same month.
Once coverage is stable, the work moves from parity to what the previous stack could not do: adding safety and operations agents on the same cameras, building an Iris detection for something specific to your process, and setting what fires on site when nobody is watching. Operators who spent the year triaging other systems' alarms are usually the fastest people in the building at spotting what to automate next.
The six that come up in every one of these conversations.
Start by pulling your own contract dates for the layer and for the video platform underneath it. Assess and pilot cost little and neither requires switching anything off, so run the comparison during the overlap and time any cutover to whichever of those two dates comes first.
No. Cobalt reads the systems already installed rather than recording video itself, so coverage during a migration is a question about the video platform underneath it. That platform keeps recording at each site until its replacement has run a full week, the cameras never move, and cutover happens site by site.
Correct, and it is the most important line on this page. Cobalt publishes that around 80% of access-control alarms need investigation rather than escalation, and that is the job it is built for. Spot AI is the video layer. If the alarm queue across access control and device health is the problem you are solving, this is not a like-for-like replacement and staying put is the right call.
It is not held by the layer, so this question belongs to whichever video platform records it today. Establish the retention and the export format for that platform in the assess phase, get anything under legal or insurance hold out before its subscription lapses, and keep that list separate from the Cobalt decision entirely.
Then two things are being decided at once and they should be separated on paper. A managed monitoring service and a video platform are different line items with different renewal dates, and a team that wants people watching overnight can keep that while changing what the cameras themselves do. Decide the video layer first, then re-scope the service against what the new platform handles without an operator.
Anyone answering this without your site list is guessing. The honest comparison is the whole stack: the video platform of record plus the layer above it, against one platform that records and reasons. Quote both over three years on the identical site list, with hardware and installation itemized.
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. Cobalt reads any RTSP or RTSPS stream and Spot AI reads any camera speaking ONVIF or RTSP, so the fleet carries across as it is. Analog cameras connect through the Intelligent Video Recorder rather than being replaced. Nothing on the wall moves and no coax is pulled.
For the video half, yes. For the rest, no, and it should be said plainly. Cobalt reasons across access control, device health and sensor alerts as well as video, and Spot AI is a video platform. Teams whose main queue is door-forced-open and badge events are solving a different problem from teams whose main queue is what happened in the yard.
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 sites you have rather than how many cameras.
Yes, and the plan depends on it. Cobalt keeps reading the systems it reads today while the IVR records the same cameras, so no comparison requires switching anything off. Nothing is retired until the site behind it has passed a full week of sign-off.
Cobalt is an intelligence layer over the systems you already run, and it says so plainly: it layers over existing infrastructure with no overhaul required. Spot AI is the video platform itself, with full-resolution video kept on the Intelligent Video Recorder in the building, agents reasoning about the frame where it is recorded, and only metadata leaving the site.