This story appears in the September 2025 issue of Utah Business. Subscribe.
“Hey, you in the red jacket. This is private property. Please move along.”
Just past midnight in a dimly lit apartment complex, a voice cuts through the silence. There’s no guard and no flashing police lights — just a camera with a view and an AI-trained security system quietly watching. Within seconds, the man in the red jacket disappears into the night. This is public safety infrastructure in 2025, and at the center is LiveView Technologies (LVT), a Utah-based company quietly transforming how businesses and cities protect people.
At first glance, LVT’s mobile security unit may look like the standard surveillance trailer you’ve seen in a retail parking lot. But rather than simply recording, the system uses real-time analysis to identify suspicious activity and respond instantly.
“The first goal of the mobile security unit is to prevent crime before it happens,” says LVT’s CTO Steve Lindsey. “We trained the AI to detect behavior that, in context, could be threatening. Like a lifeguard blowing a whistle at a pool, it calls them out on what they’re wearing, what they’re doing, and tells them to stop.”
Looking for red actors in a sea of green
Imagine a park or festival where there are hundreds of people, most of whom are doing nothing wrong. The LVT system watches this “sea of green actors” for “red actor” behavior that could indicate wrongdoing. Is someone pacing between cars? Approaching vehicles and peering into windows? Loitering around an ATM for an extended period of time?
Once the system identifies a threat, it doesn’t just play a canned message. It creates a custom voice command using generative AI. The voice is modeled after LVT CEO Ryan Porter. “We actually did it as a joke,” Lindsey says, “but it sounds so compelling that we decided to keep it.”
Avoiding false positives
With technology like this, mistakes can have serious consequences. What happens if the system misidentifies a wallet for a gun, and police overreact because they expected to find someone armed and dangerous?
“There’s a human in the loop filtering before it’s escalated to where the police get involved,” Lindsey says. “For instance, if I’m in a place where someone brandishes a weapon, it’s immediately escalated to the entity who hired our services. We provide enough actionable intelligence that they can look at the situation and make a final assessment.”
LVT also made a deliberate decision to avoid training on data that could lead to discrimination or stereotyping.
“We try to stay away from vulnerable technologies like facial recognition,” Lindsey says. “We’re more interested in observing a scene in the context of what’s happening than in personally identifiable information.”
What’s next?
LVT has grown quickly in recent years, and you can find mobile security units monitoring everything from critical infrastructure to retail, solar fields, construction sites and even government facilities. But despite its success, the LVT team remains focused on bettering the technology.
“You’re never going to prevent crime 100 percent, but we see ourselves continually improving the prevention side of things: … better AI, better intelligence, more accurate decisions. … The best case scenario is we put a unit somewhere, and nothing happens.”
Because sometimes, just watching is enough to stop crime from occurring at all.
