Davista takes security AI beyond video surveillance

Oct. 21, 2020
Technology originally developed for the government now being leveraged in commercial applications

The terms artificial intelligence (AI) and machine learning have almost become synonymous in today’s security industry with intelligent video analysis solutions. However, while many of these technologies are relegated to deriving some type of actionable data from a camera feed, few have tried to apply that to the entirety of an organization’s security operations. That is the goal of Davista, which has developed an AI platform that is designed to key in on the data generated by a multitude of security devices and systems.   

According to Scott Sieracki, the company’s Chief Sales and Marketing Officer, Davista is focused on the organizations that operate some semblance of a security operations center – be it a large, global security operations center (GSOC) environment or a two- or three-person department that still has the responsibility for managing the security technologies throughout a company.

“What we do with our AI engine is collect information from typically what are very disparate silos of information. If you start with the conventional stuff in let’s say the average security operations center, we can pull data from physical access control systems and video management systems,” explains Sieracki, who has also served in executive roles with access control firms Identiv and Viscount. “When we touch video management systems, we’re not touching the video; we’re actually touching the metadata. We want the data about the video. We can also pull from intrusion detection, incident management – albeit we have our own incident management module in the application – as well as from security workforce management applications, which we have a module for as well.”

Once all of this data is collected, Sieracki says they can start to build a bridge with other parts of the organizations that might have insights relevant to security, such as logical security, logins to voice-over-IP (VoIP) systems, and conference room management systems, with the goal of aggregating all of the things that might be relevant to a security incident.

“What we’re doing with that data is we’re giving the customer, probably for the first time in their environment, an analytical view of the trends that are taking place in their business,” Sieracki adds. “There is part of us that can be historical; if you look at what goes on in the security environment, whether it is the conventional corporate security environment that ASIS members belong to or even law enforcement, it is typically a tactical environment and a reactive environment.

“We’re aggregating that data, but what we’re starting to do is make what we call that sort of ‘Minority Report’ dashboard for the end-user,” he continues. “We’re starting to take the trends and analysis coming out of as much historical data we can get so that we can start to prescribe proactive steps for the security environment to take to start to mitigate, remove or lessen risk, impact or events.”   

Government Origins

Abish Malik, Davista’s Founder and Chief Technology Officer, says the company got its start in one of the Department of Homeland Security’s Centers of Excellence, which are essentially thinktanks at universities around the nation that bring together security and law enforcement officials with technology developers to try and solve problems that impact national security. The center that Malik was working at just so happened to be tasked with the challenge of handling data generated by DHS and other agencies.

“We had the privilege of working for the Coast Guard, TSA, intelligence agencies, numerous police departments across the U.S., USCIS, and CBP. The beauty of working at the center was we got to work with so many different facets of this enterprise-level organization and they started the data movement – the government started the data movement,” Malik says. “Our mandate was to help our customers utilize data that exists out there that they have. Instead of the data just sitting there, how do we operationalize the data from a short-term, day-to-day perspective to a long-term, strategic perspective as well? What we eventually built for them, that is now what we call ‘Heimdall.’”    

A New Paradigm

Malik added that Heimdall, which can adapt to multiple problems at scale, is creating a new segment in the security market.

“What makes this so relevant and unique is that for the first time, there is an AI story that helps customers get out of their tactical, reactive environment,” Sieracki adds. “We’re not a PSIM, so we’re not looking to bring in alarms and video and other subsystems a PSIM looks at and then announces that something has already happened. We’re looking to connect to those systems, listen to the data, analyze the data, and then start to project from that data ways the business can stay ahead of things.”

For example, Sieracki says the company is currently working with a customer in the airport environment and their platform can help security spot anomalies in things like changes in employee behavior.  

“They’ve got an issue with their own airport employees, people who are badged who normally come in through exit A all day long and then TSA moves a random checkpoint to entrance A and now certain employees will stop using entrance A and will go find entrance B to come through so they can avoid a security checkpoint,” he says. “In our system, we have the ability to see that as an anomaly almost instantaneously and we have the ability, for example, to call that to a security manager’s attention that, ‘hey, you’ve got access control behavior that is non-typical based upon the historical data that we’ve looked at over the course of the last 30 days, 60 days or whatever it might be. This is similar to insider threats where we can tell an organization that somebody’s access entitlements are operating at a different behavior than they were before said day.”

Preventing the Spread of COVID

Davista has also adapted Heimdall for mitigating the spread of the coronavirus as the company recently announced a new social distancing platform designed to help organizations facilitate safe return to work strategies.

“Instead of it being an app that just sits in the hands of a guard or the management folks, we have a version of that app that’s on John Q or Suzy Q employee’s phone and we’re able to a) track people all day long and b) let each other know when someone has violated social distancing requirements,” he explains. “From the fact that we’re helping from a proactive level to keep people away from each other, we also have a highly intuitive COVID contact tracing window. We can tell administrators, ‘hey, you’ve got a handful of people that are potentially violating the social distancing requirements.’” 

And while AI may not be new to the security industry and does not necessarily intimidate the average end-user, consultant or integrator anymore, Sieracki says that they still have to educate the broader market that this technology can be used for more than just video analytics.

“AI is not a brand-new concept to our industry. The cool part for us is we get to ride or piggyback on the fact that AI has been introduced into our domain,” he says. “What we have to do from a messaging perspective as well as an education perspective, is help people understand there is a bigger picture to AI in security operations.”

Joel Griffin is the Editor-in-Chief of SecurityInfoWatch.com and a veteran security journalist. You can reach him at [email protected].