A Florida tech firm says it can help pull Los Angeles’ overloaded 911 system back from the brink by letting artificial intelligence handle routine complaints.

Public safety technology company Aurelian has developed an AI-powered call taker that answers police non-emergency phone lines, takes reports and routes complaints.

The system handles everything from barking dog complaints and parking disputes to abandoned vehicles, lost property reports and suspicious activity calls.

The pitch comes as Los Angeles continues battling chronic dispatcher shortages and slow emergency response times.

A March report to the City Council found LAPD answered just 57.43% of 911 calls within California’s 15-second standard in 2024, far below the state’s benchmark requiring 90% of emergency calls to be answered within that timeframe.

Staffing remains a major challenge.

The California Post previously reported LAPD hired 144 dispatcher trainees in 2024 but only 56 in 2025 while losing 75 operators during roughly the same period.

City officials have said about 100 operators must be on duty across a 24-hour period just to meet minimum staffing requirements.

According to Aurelian CEO Max Keenan, roughly 70% of calls entering a typical emergency communications center are not emergencies at all.

“They’re barking dogs. Parking disputes. Noise complaints. Lost bicycles. Requests for city services. The same people answer both calls,” Keenan told The California Post.

“The analogy I use now is you basically train your team as Navy SEALs and you use them as mall cops. It is like the greatest misallocation of labor imaginable.”

That reality led the company to build Ava, software designed specifically for the 10-digit non-emergency phone numbers operated by police departments and emergency communications centers.

Download The California Post App, follow us on social, and subscribe to our newsletters

Keenan said the scale of the problem became clear while visiting dispatch centers around the country.

One center told him it handled about 200,000 emergency calls a year.

“I did the math and I was like, 200,000 calls a year, for 100 people, that doesn’t seem that bad,” Keenan said. “And she said to me, ‘Oh, we take a million non-emergency calls.’”

According to the company, Ava automates about 74% of non-emergency calls and gives dispatchers back roughly three hours of their workday.

Keenan stressed that the technology never answers 911 calls.

“We only take calls that are dialed into the 10-digit non-emergency line,” he said. “We are not taking any calls that touch 911.”

Still, the system continuously screens calls for emergencies and immediately transfers callers to dispatchers when necessary.

In one recent case, a caller reported a woman lying on railroad tracks through a non-emergency line. The AI recognized the danger and escalated the call to a dispatcher.

“About 5% of the calls in the non-emergency line are emergencies,” Keenan said. “Being able to understand what’s happening in the situation and escalate that call to a 911 call taker is incredibly important.”

Agencies using the system typically automate between 60% and 80% of their non-emergency calls, according to Keenan.

One department that previously fielded about 80,000 non-emergency calls annually now has dispatchers personally answer only 15,000 to 20,000 calls, with Ava handling the rest.

“It’s one of the hardest jobs in the world,” Keenan said of emergency dispatchers. “It’s not paid well enough for the difficulty, for the amount of skill that it requires, and in some ways there’s some level of trauma.”

He said staffing shortages have plagued emergency communications centers for decades.

“Staffing has been a massive issue. It’s been a crisis for decades.”

Keenan says the answer is not simply hiring more people.

“You can’t hire your way out of a retention issue and oftentimes you can’t really pay your way out of it either,” he said. “You have to be more intentional about how you make the quality of life better.”

Keenan said the company serves more than 50 public safety agencies and processes hundreds of thousands of calls every month. And unlike many government technology projects, he said agencies can deploy the system quickly.

“Oftentimes we’re going live with a customer in less than 10 weeks from a contract signed to go live,” Keenan said. “Usually day one, we’re automating 60% of their calls.”