Introducing Smarter, Faster Verification for Video Monitoring
We’re constantly thinking about how we can deliver more value to our partners, which in turn increases reliability for their end users. That’s why we’re so excited to share new updates to our Verify API and video monitoring solution that do just that — help security teams move faster, reduce false alarms, and improve the experience for their end users.
This is achieved through two new features added to the Verify API: AI Person Filtering and Advanced Verification. Together, these result in fewer disruptions, fewer unnecessary police dispatches, and faster resolution when it matters most. Here’s what you can expect from these new features, and why they make such a difference for your video monitoring solution.
Verify API and Video Monitoring Overview
Our Verify API pairs powerful analytics with the expertise of Noonlight agents to detect and verify real emergencies. Once a potential threat has been detected by a camera (like a person in a construction zone after hours or someone loitering near an empty store) the footage is pushed to a Noonlight agent to verify the threat.
If it turns out to be a shadow or foreman there to collect his forgotten keys, it’s business as usual without an alert. If it appears to be a break-in or bad actor, agents trigger an alarm to dispatch first responders and notify users via SMS within seconds.
These capabilities are especially impactful for both commercial and residential security solutions, from protecting construction sites, retail storefronts, and office buildings after hours to giving homeowners added confidence that real threats are identified and addressed quickly without unnecessary disruption.
New Feature: AI Person Filtering
To identify real threats quickly and accurately, it’s vital for only pertinent footage to be sent to agents. Our new proprietary AI Person Filtering model does just that by filtering and dismissing video events where no people are detected.
Our industry-leading model scans and processes hundreds of frames per video clip in order to minimize any chance of missing even brief appearances by a person, resulting in 99% recall and 97% precision. Other solutions in the space skip frames, leading to a higher miss risk.
This process reduces motion noise and ensures that monitoring agents focus is only on real events, not non-actionable footage, all while lowering operational cost.
New Feature: Advanced Verification
99.4% of events flagged by video verification end up being low-level threats that don’t require contacting emergency response. To reduce unnecessary alarms and false dispatch, we enhanced our existing verification workflow with Advanced Verification.
Advanced Verification provides agents with additional context to make more informed decisions after a triggering event; including access to live feed, additional footage, and talkdown deterrence in real time—all before an alarm is created. As an end-to-end monitoring solution, agents will monitor the situation until resolution or trigger an alarm if necessary. These additional features enable faster intervention, improved response times, and a better experience for end users, all without compromising safety or security.
Additional Talkdown Capabilities
Real-time voice communication allows agents to move beyond simple assessment into active engagement. In many cases, talkdown is enough to defuse tension, deter unwanted behavior, and resolve incidents before they require further escalation. Early data indicates that Advanced Verification with talkdown:
- Reduces total false alarms by 45%
- In 37% of cases, the person left the scene on their own
- In 8% of cases, the agents were able to determine the person was an authorized employee
This exemplifies how direct voice intervention can prevent damage or theft before it occurs—reducing unnecessary alarms while ensuring real threats aren’t overlooked. In many cases, that early engagement eliminates the need for escalation altogether. As a result, emergency dispatch is no longer the default response. Agents can often resolve situations directly, with dispatch reserved for cases that truly require it.
Ready to See for Yourself?
Our Verify API updates further our mission to support and comfort people so they can live freely — now with even more accuracy and context to ensure your end users are supported through real emergencies and receive the fastest resolution possible.
Noonlight President John Tassone said it best: “What we’ve built is a smarter way to deliver the right context to agents at exactly the right moment — so they can filter out the noise, act decisively on real situations, and prevent emergencies from escalating wherever possible.”
.svg){kind=small}
