Noonlight adds new upgrades to its Verify API, introducing AI-based person filtering and enhanced verification workflows designed to reduce false alarms and improve response efficiency.
As video surveillance systems become more advanced, the volume of alarms generated by motion detection and analytics continues to rise. For security providers and monitoring centers, separating real threats from routine activity remains a persistent challenge.
US-based Noonlight is addressing this issue with new upgrades to its Verify API, introducing AI-based person filtering and enhanced verification workflows designed to reduce false alarms and improve response efficiency.
The company’s approach reflects a growing trend in the industry: combining cloud-based analytics with human verification to improve decision-making without overwhelming operators.
API-based architecture supports flexible integration
A key aspect of Noonlight’s offering is its API-first design, which allows it to integrate with a wide range of existing video systems.
“Verify is an API-based platform, which means it is both camera-agnostic and VMS-agnostic,” said John Tassone, President of Noonlight. “Any security provider that can generate an API call to our platform when a camera is activated, and include a video attachment (multiple formats are supported, such as MP4 or M3U8), can integrate with Verify.”
This approach allows security providers and integrators to deploy the solution without replacing existing cameras or video management systems.
Because the platform relies on standard API calls and video payloads, the integration process is described as relatively straightforward.
“Because of this flexible architecture, the integration process is typically straightforward for security system integrators with API capabilities,” Tassone said. “Once an event trigger and video payload can be sent via API, the system can begin leveraging Verify.”
The company’s current partners include DTiQ, Wyze, Envysion, Solink, Omnilert and TrueLook, indicating adoption across different verticals and deployment types.
For integrators, this suggests that Verify can be layered onto existing systems rather than deployed as a standalone solution.
Cloud-based AI aims to improve detection accuracy
With many modern cameras already offering built-in analytics, Noonlight positions its AI person filtering capability as a more advanced alternative.
“Many cameras and VMS platforms run analytics directly on the edge device, which means they are constrained by the limited processing power available on the camera itself,” Tassone said. “This often results in lower accuracy and a higher rate of false positives.”
By contrast, Noonlight processes video events in the cloud, where more computational resources are available.
“Noonlight’s person detection runs in the cloud, allowing us to leverage significantly more powerful AI models that deliver faster processing and higher accuracy,” Tassone said. “By removing the hardware limitations of edge devices, we can apply more advanced detection and filtering to ensure that only relevant events are surfaced.”
This cloud-based approach reflects a broader industry shift, where vendors are increasingly using centralized processing to improve analytics performance.
At the same time, the model raises familiar considerations around bandwidth, latency and data handling, which integrators must balance against the potential gains in accuracy.
Streamlined workflows for monitoring operations
In addition to improving detection, Noonlight’s Verify platform is designed to change how video events are handled operationally.
Unlike traditional setups that rely on third-party monitoring centers, Verify events are reviewed by Noonlight-authorized agents.
“Verify is not offered directly to third-party monitoring centers. Instead, all events are handled by Noonlight-authorized agents,” Tassone said.
The workflow itself is built around reviewing short video clips triggered by events, rather than continuously monitoring live feeds.
“Verify delivers a simple, focused event review experience centered around short video clips tied to specific triggers,” Tassone said.
This design is intended to reduce operator fatigue and improve accuracy. By reducing cognitive load, the platform aims to enable faster and more consistent decision-making.
“This approach allows agents to quickly assess incidents with minimal cognitive load,
improving review speed and accuracy,” Tassone said.
For security providers, this could translate into more efficient monitoring operations, particularly as event volumes continue to grow.
Data shows significant reduction in false alarms
Noonlight reports that its combined AI and human verification approach is already delivering measurable results in commercial deployments.
“Across our commercial partners, Verify has demonstrated meaningful improvements in both false alarm reduction and response efficiency,” Tassone said. “For partners using our newer Verify capabilities, we’ve seen a 45% reduction in false alarms.”
The addition of intervention tools, such as Talkdown, allows agents to interact directly with individuals on-site.
“When agents use intervention tools such as Talkdown, individuals leave the scene voluntarily 37% of the time,” Tassone said. “In an additional 8% of cases, extended footage allows agents to determine the person is an authorized employee—preventing unnecessary police dispatch while ensuring legitimate threats are not missed.”
Looking at broader operational data, the company highlights the scale of event handling.
“In 2025, across all Noonlight partners using Verify, Noonlight agents reviewed 7.5M+ events,” Tassone said. “99.3% of false alarms were filtered out before alarm creation and police dispatch,” he added.
The average verified response time was reported at 29 seconds, with over 150 confirmed arrests linked to verified incidents.
For integrators and end users, these figures highlight the potential impact of improved alarm verification on both operational efficiency and public safety.
Identifying the root causes of false alarms
The data also provides insight into why false alarms occur in commercial environments.
“69% – Employees performing routine activity; 18% – Person or vehicle in a public area; 11% – No person visible in the clip; 2% – Other causes,” Tassone said.
These findings underline the limitations of traditional motion detection and even some forms of video analytics.
For integrators, this reinforces the need for solutions that can provide better context and filtering before alarms are escalated.
Subscription model supports scalable deployments
In terms of business model, Noonlight is positioning Verify as a predictable, subscription-based service.
“Verify is offered as a monthly subscription based on the number of deployed camera endpoints (nodes),” Tassone said.
Each subscription includes a pool of human-reviewed events that can be shared across the deployment.
“Each subscription tier includes a pool of human-reviewed verification events that can be used across the entire account, allowing partners to flexibly allocate usage across their deployment rather than being constrained by limits on individual devices,” he added.
The company also highlights a departure from per-event pricing models.
“With Verify, AI-powered event filtering is included with unlimited events as part of the subscription, so partners don’t have to worry about volume-based overages as their deployments grow,” Tassone said.
For integrators, this approach could simplify pricing and make it easier to scale deployments without unpredictable costs.
Hybrid models gain traction in video monitoring
Noonlight’s latest upgrade highlights a broader shift in the security industry toward hybrid monitoring models that combine automation with human oversight.
As video analytics continue to improve, the challenge is no longer just detecting events, but determining which events matter.
By using AI to filter large volumes of video data and human agents to verify context and take action, solutions like Verify aim to bridge this gap.
For security systems integrators, the key considerations will be how such platforms integrate with existing systems, how they impact operational workflows, and how they can be packaged as part of a broader service offering.
As the volume of video data continues to grow, solutions that can reduce noise while maintaining reliability are likely to play an increasingly important role in the evolution of video monitoring.