What HiveWatch's own pages say its GSOC operating system does not do, and what it depends on the systems beneath it to do, written for security operations leaders running due diligence. Every constraint below carries the documentation behind it, the operational impact, and the workaround.
Every constraint below comes from HiveWatch's own current published pages, checked on August 17, 2026. Where HiveWatch documents nothing, this page says so rather than guessing, because an absence on a public page is not a finding about a product. Spot AI sells a competing platform, so nothing here rests on an anonymous source or an aggregated user rating, and two claims are corrected on this page. HiveWatch is not vague about what it connects to: it names the video and access platforms on its own site. And it is no longer only a platform for teams that already run a command center, because it now sells HiveWatch GSOC, described as 24/7/365 coverage from its own expert team running on the HiveWatch platform.
The systems it connects to are named in public and the operator workflow is published in detail. Whether an alarm was ever raised, and what happens on site when one is, belong to somebody else.
A constraint list is only useful next to an honest account of the product.
Genetec, Milestone and Verkada on video, OnGuard, LenelS2 and Brivo on access, plus Ontic and AlertMedia, with Johnson Controls, Cisco, ExacqVision, Motorola and Honeywell in the wider partner network. A security architect can hold a real stack against a real list before a call rather than asking whether an integration exists and being told it can be built.
HiveWatch states there is no need to rip-and-replace and describes itself as the intelligence layer over what is already installed, with the reasoning in one line of its own: you can only detect what you can connect. For an estate carrying three video platforms from three acquisitions, that is a materially cheaper path than a consolidation project.
The published figures are 22 seconds average incident resolution time and around 90% reduction in false alarms, with a Fortune 500 gaming company reducing false positives by more than 60% in a few months while saving millions in rip-and-replace projects, a law firm at $500K in savings and a hyper-growth technology company at $28M. Those are operational metrics rather than adjectives, and they are attached to a named customer profile.
Each one is a consequence of how the platform is designed rather than a defect. What matters is whether it collides with your starting point.
HiveWatch describes itself as the intelligence layer an existing stack has been missing, and states that every alarm type, from access control and video through employee quick reports, weather and fire, flows into one operator workflow. Tailgate detection is the one detection named as HiveWatch's own on the pages this page read. Its own line states the dependency plainly: you can only detect what you can connect.
If the video platform underneath does not raise an event, there is nothing to route. A campus whose cameras record faithfully and notice nothing will get a faster, tidier queue of the same incomplete alarms, and the incidents nobody caught stay uncaught.
Take one week of real alarms and split them into two piles: real events handled too slowly, and events that never fired or fired for nothing. If the first pile is bigger, HiveWatch is aimed at your problem. If the second is bigger, fix detection first and orchestrate afterwards.
The documented output is an operator workflow: one queue, Alert Automation filtering what reaches it, Incident Management running the response, an AI Operator working alongside people in what HiveWatch calls human-in-the-loop operations, with guard tour management, dispatch SLAs and mobile incident logging around it. Active deterrence on site is not publicly specified. The pages describe what reaches a person, not what happens at the fence.
For a corporate campus with staffed operations that is the design working as intended. For a distribution site whose real requirement is that somebody in the yard at 03:00 leaves rather than that the event is well documented, the response still waits for a person to be watching and available.
HiveWatch has its own answer to the staffing half, and it is worth taking: HiveWatch GSOC is published as 24/7/365 coverage from its expert team running on the platform, aimed at teams that are under-resourced or do not want to build a command center. Price that alongside the software rather than assuming the queue covers unstaffed hours, and keep the response capability itself on the platform that owns the cameras, asking which named integrations can fire a speaker or a strobe today.
HiveWatch states there is no need to rip-and-replace and positions itself over the systems already installed, naming Genetec, Milestone and Verkada on video and OnGuard, LenelS2 and Brivo on access. Its guard tooling is published in the same terms: guard tour management, dispatch SLAs, mobile incident logging and performance reporting so guard spend is defensible.
The three-year model has three live lines rather than one: the platforms already in place with their renewals, HiveWatch above them, and the guard program it makes defensible. Nothing in the published material retires any of the three, which is honest and does mean the total goes up before the operator hours come down.
Model all three lines side by side, then agree the metric that has to move before signing: alarms per operator, time to resolve, or false positives per site. HiveWatch publishes numbers in exactly those units, so ask for the target in the same units and put it in the contract review.
No rollout timeline, hosting model or integration effort per system appears on the pages this page read. The documentation covers which systems can be connected, not what connecting one costs or how long it takes. That is an absence on a public page rather than a statement about the work involved.
Connecting six consoles across two acquisitions is exactly the kind of project whose cost lives in the effort rather than the license, and it is invisible until a scoping call. A program approved on a license figure can absorb its first quarter in integration work nobody sized.
Ask for a statement of work per connected system, with the version of each platform named, and a dated rollout plan for the first three. Make each connector a separate line rather than a bundled onboarding fee, so a system that turns out to be difficult is visible rather than absorbed.
The same 28 million dollar saving is presented on the home page as an average saved yearly for one customer, on the why-HiveWatch page as saved over three years, and on the customer stories page as long-term savings. The false-alarm figure moves too, at 90% on one page and 91% on another. Separately, no SOC 2, ISO or equivalent statement appears on the pages this page read, next to a product that aggregates alarm data from every security system in the estate.
A number that lands in a board paper needs one meaning, and a reviewer who spots the discrepancy will discount the whole business case rather than the sentence. The certification question is the one a security committee always asks of a platform holding that much aggregated data.
Ask which of the three readings is correct and for the customer's own before and after, then build the case from your own alarm volume rather than a reference figure. Request the certification pack early, naming the documents you need: the current SOC 2 report, the data flow description and the retention terms for alarm data.
The same five constraints in one view, sized to paste into an evaluation document.
Swipe the table sideways to see every column.
HiveWatch data comes from HiveWatch's own site and GSOC operating system pages, checked on August 17, 2026. Gaps are marked as not publicly specified.
If there are operators, six consoles and an alarm volume nobody can work through, HiveWatch is built for precisely that problem and names the systems it plugs into before you ask. Most of the constraints above stop applying when the alarms are real and the queue is the bottleneck, because that is the job the product was designed for, and the published outcomes are in the right units for it: resolution time, false positives, alarm capacity per operator and the reporting cycle.
Spot AI sits one layer down, which is why its constraint list looks different. Rather than routing what other systems raise, it changes what gets raised: 15+ pre-trained Video AI Agents cover vehicle break-in, fire, cash register theft, after-hours intrusion, personal protective equipment, forklift near-miss, falls and crowding in hazard zones, and Iris builds a custom detection in natural conversation in about eight minutes. The detection quality is the platform's own rather than inherited from whatever was installed in 2018.
The response is on the platform too. AI Security Guard reads the context and then answers within seconds through talk down, strobes or horns on standard speakers already on site, so an unstaffed yard at 03:00 has a documented path that does not begin with a person noticing a queue. Any ONVIF or RTSP IP camera connects as it is, legacy analog comes in through the Intelligent Video Recorder, full-resolution video stays on the IVR in the building, and only event metadata crosses the network.
Better routing of poor detections is still poor detections. Score the two against the same week of real alarms rather than against a feature list.
None of that makes Spot AI the right answer for every buyer. HiveWatch publishes an integration list by vendor name that Spot AI does not match: OnGuard, LenelS2, Brivo, Ontic and AlertMedia are named on its own pages, where Spot AI reaches other systems through open APIs, webhooks and an MCP endpoint. HiveWatch is also built to be the single queue over platforms a large enterprise has no intention of replacing, and Spot AI is not that: it is a video AI platform, so an estate committed to three video management systems across three regions is being asked to consolidate rather than to orchestrate. Access control and sensors arrive through integrations. A constraint list earns its keep by holding each platform's shape against your starting point.
A live pilot on your cameras answers in a week what a spec sheet cannot.
Customer-reported outcomes from named Spot AI customers.
Cambridge City cut footage search from two hours to 30 seconds after consolidating seven municipal locations on one platform.
Don Franklin Family of Dealerships had incident footage on responding officers' phones within four minutes of an alarm, and reported five recoveries within the hour.
Unique Industries covers more than a million square feet with a three-person safety team, catching near misses and falls on the cameras already installed.
"With Spot AI, these facilities are more secure and better watched than the ones we staff traditionally, where we pay for a full-time person on-site."
Five, all documented: it improves what happens to an alarm rather than whether one was raised, with tailgate detection the one detection named as its own; the documented output is an operator queue, so on-site deterrence is not publicly specified; it sits over systems that keep their own licenses and over a guard program it makes defensible; the deployment model and the effort per connected system are not publicly specified; and one published saving appears two ways, with no certification statement beside it.
Through your video platform, and it names which ones: Genetec, Milestone and Verkada on video, with OnGuard, LenelS2 and Brivo on access control and Ontic and AlertMedia alongside them. HiveWatch connects to the system that owns the cameras rather than to the cameras themselves, and states there is no need to rip-and-replace. Check your own stack against that list before the first call, including the version of each platform you actually run.
No, and it does not claim to. HiveWatch describes itself as the intelligence layer over the systems already installed, and the alarms it works with arrive from the video and access platforms it connects to. That means the quality of what reaches an operator is still set by the platforms underneath it, which is the single most important thing to understand before shortlisting it. Tailgate detection is the one detection named as HiveWatch's own.
Three documents and one clarification. The current SOC 2 or equivalent report, since no certification statement appears on the published pages next to a product that aggregates alarm data from every system in the estate; a data flow description with the retention terms for that alarm data; a statement of work per connected system; and confirmation of which reading of the published 28 million dollar saving is correct, since the pages state it as a yearly average, as a three-year total and as long-term savings.
It depends which half of the problem is yours. If alarms are being missed rather than mishandled, or there is nobody watching a queue overnight, a camera-agnostic platform such as Spot AI fits: 15+ pre-trained Video AI Agents run across the cameras already installed, AI Security Guard answers through talk down, strobes and horns on standard speakers, and full-resolution video stays on the Intelligent Video Recorder on site. If the alarms are real and the queue is the bottleneck, HiveWatch is aimed at exactly that, and for a team with no command center of its own it now sells HiveWatch GSOC, published as 24/7/365 coverage from its own expert team running on the platform.