Video AgentsAll tags
Agentic VMS

Brivo's Eeva turns a sentence into a camera rule

The Eagle Eye VMS now takes a plain-English description of what to watch for, such as a worker without a safety vest, and starts sending notifications. No new hardware, and one integrator has it guarding EV charging cables.

By Daniel Reyes · 2 min read
A construction worker in a hard hat and high-visibility vest walking away from the camera on a paved plaza
The example prompt in Brivo's announcement is a single sentence: identify when a worker is not wearing a safety vest. Photo: Lei Hwang / Unsplash

Brivo announced Eeva on March 17, 2026, and the notable thing about the announcement is how ordinary it makes the idea sound. The company's post, published without a byline on the Brivo blog and timed for ISC West in Las Vegas the following week, presents Eeva as a feature of the Brivo Eagle Eye VMS. You type what you want the cameras to notice. Eeva watches for it.

The example prompt in the post is "Identify when a worker is not wearing a safety vest." Once that sentence is entered, the system begins monitoring the selected cameras and sends notifications when it finds a match. Automated actions can be attached to the same trigger. Brivo says the underlying model is a vision-language model, that the agent takes the context of the scene into account, and that it runs on whatever cameras a customer already has connected to the VMS. There is no appliance to buy, and a deployment can cover thousands of cameras at once.

The employee comparison

Dean Drako, Brivo's chief executive and the founder of Eagle Eye Networks before the two companies combined, described the agent in the terms the industry has settled on. "Eeva is like an ideal employee working 24/7 all year round—someone who's reliable, diligent, takes action independently and is always on the job," he said.

The more useful material is further down, where two integrators describe what they used it for. Lee Kerry of Kings Secure Technologies had it watching electric vehicle charging stations, where cable theft is a recurring loss. "Eeva's AI can distinguish when a person or vehicle is present at the EV station from when a person is attempting to remove the cable from the protected zone," Kerry said. A motion rule fires for the customer plugging in and for the thief unplugging alike, and Kerry only wanted the second.

Edward-Oliver Olafsen of Enter Security said setup was straightforward and the detections have held up.

Brivo lists other early uses: illegal dumping at managed properties, glove compliance in restaurant kitchens, safety vests on factory floors. Each is a rule a site manager could state in a single sentence and has never been able to enforce without a person watching.

I would want to know how the agent behaves when it is wrong, how a customer tunes it after a false positive, and whether Eeva writes anything up after an event. The post claims detection and notification. Investigation and documentation, which competitors have started to bundle under the same word, go unmentioned. Brivo's VMS already sits in a large installed base, though, and for a lot of those customers a prompt box on top of the cameras they already have may be all they wanted.

Read the original at Brivo.