Veritone continues to see growth opportunities for its aiWARE artificial intelligence (AI) operating system among existing media customers and clients in new business sectors including public safety, according to Veritone CEO and chairman Chad Steelberg.
“The biggest market opportunity that we are now tackling” is in the public safety area, specifically working with the federal government, he said Aug. 8 at the Oppenheimer Technology, Internet & Communications Conference in Boston.
Veritone recently went through the process of becoming certified to provide services to the federal government through the Federal Risk and Authorization Management Program (FedRAMP), he noted. It completed the certification process to provide services to the Department of Justice (DOJ) and other U.S. government agencies, he told the conference.
Veritone now has paid proof of concept (PoC) deals for the identification of known offenders through facial recognition and to support the DOJ and local district attorneys in the processing of audio and video evidence received from body and dash cameras, which are increasingly being used, he noted.
Already, Veritone has been able to “scale across the majority of the titans of the media industry today,” he told the conference. “There’s literally probably not a media company in the United States that isn’t either on an annual contract basis with us or in a long-term PoC,” he said, adding: “Those customers range from infrastructure companies in the media industry to the classic broadcasters” including iHeartMedia, CBS and Disney/ESPN “to now actually the content owners themselves,” including sports leagues that it has NDAs with and can’t disclose.
Citing Veritone’s performance in the first quarter (ended March 31), he pointed to its growth in Software-as-a-Service (SaaS) as the “big story.” Total revenue jumped 41% from a year ago to a record $4.4 million, with SaaS licensing revenue from the aiWARE platform soaring 506% to a record $1.3 million.
Veritone’s SaaS revenue growth reflects the success the company started to have with its “land and expand” strategy of making it easy and cost-effective for customers to deploy AI to enable solutions to their immediate business needs, and then increase business with those customers as they realized the strength of Veritone’s platform and created additional use cases, according to Steelberg.
As an example, he noted that Veritone started offering ad verification to ESPN and has since grown its business with that client. Initially, ESPN had only a monthly contract with Veritone for broadcast radio, starting in 2016. But the pact became an annual one that added TV simulcasts and podcasts later that year and expanded internationally in 2018, according to a slide he showed during his presentation. Revenue from ESPN soared 325% in less than two years, according to Veritone.
From its start in 2014, Veritone’s “mission” was to “deploy artificial intelligence to every person and every organization on the planet,” Steelberg said, adding: “Our strategy for doing that really [revolves] around one core challenge that we all have, which is, in the cloud today, over 80 percent of every byte stored is of the form of unstructured data, principally audio and video. That information is completely locked from computer understanding and it requires humans to translate it to structured data for it to be analyzed.”
But, thanks to AI and machine learning, “machines can now directly analyze that information and unlock its value,” he said, adding Veritone’s aiWARE AI platform was “designed specifically for the enterprise to do that at scale, both in the cloud” and on-premises. “That pie of all the data stored in the cloud is growing, measured today in zettabytes… is effectively doubling every 18 to 24 months [and] we expect that to continue for the next decade,” he said.
When it started, Veritone could only find five commercially available modern AI engines, including Microsoft’s, he said. But Veritone forecast there would be a “rapid explosion” and, today, we’re tracking over 10,000 companies in the marketplace developing cognitive engines in over 60 classes of cognition,” he said, adding: “We’re expecting the number of engines to go from 10,000 to millions in the next 10 years.”