SVG Sit-Down: Veritone’s Sean King on the Power of Mining Video, Audio Data

The company’s Data Refinery offers users total control and governance over data assets

Lots of companies are talking about AI and its role in the sports-media landscape, but Veritone is one of the leaders when it comes to deals, with organizations like the Masters, the USTA, and the NCAA on board as clients. Veritone Chief Revenue Officer/GM, Commercial, Sean King leads all sales and marketing activities for the organization, in addition to overseeing the Commercial division, including SaaS technologies and managed services. He sat down with SVG to discuss key market trends, the Veritone tech stack, and much more.

Veritone’s Sean King is looking forward to a busy 2026 with the Winter Olympics and the World Cup.

From a Veritone perspective, what have been some of the key milestones of the past year?
Our Data Refinery, which is where we’re helping all our customers in both the commercial and public sector, [gain] total control and governance over their data assets, specifically their unstructured data, all their audio and video.

As we’re ingesting all the footage — either onsite or remote — we’re taking that content, making it structured, making it searchable, and helping them have the necessary governance. If they want to take their assets and make them available for search and discovery to be licensed for secondary and tertiary use cases or if they want to use that data to build an internal workflow or model or even make that data licensable to other companies, it allows us to give them a view into all their data, structured or unstructured, and the necessary tools to govern, to monetize, or to manage it.

In terms of data, there are the stats — lineups, things like that — that are input into video files. Then there’s the audio track with announcers, captioning information, and, of course, facial recognition and object tracking. When you look at that ocean of data, what kind of stuff boils up and makes the most sense?
Specifically, we look at everything that’s in the audio and video. We look at third-party data, from Stats Perform or Sports Radar, that we can time-correlate to what’s taking place. But, for us, it’s everything that’s taking place on screen, whether it be face, logo, or objects.

You could say, “Okay, let me find an image of Tiger Woods in a red Nike shirt and let there be an image of a Rolex in the background. We search through all those [images] and find what meets that multivariant search criteria. It’s about giving that frame-by-frame annotation of everything on the screen or in the recording, and, if someone wants to marry that with those third-party data sources, you have the total flexibility to do so.

One of the challenges, obviously, is to manage archives and begin the process of digitizing things. What is your advice for someone who hasn’t begun the journey? Even colleges, universities, and high schools are looking at what they should be doing. What are your tips?
It’s a great question, and I think you’ve got to look at it two-fold. I always like to give the analogy of driving your car. The first thing you have to do is pay attention to what’s in the windshield in front of you. Everything that’s coming at me: how am I going to make sure that, as I’m going forward, I’m capturing everything? Making sure that I have all the control, all the data, and all the monetization opportunities — that’s step one.

Step two is, Okay, now let me look in my rearview mirror and let me see all the places that I’ve sat before. The catch-up, especially the digitization, can be a time-consuming and costly endeavor. Start with the high-value, high-touch ones like the big historical moments. Once you start there, you can see the monetization opportunities and let those opportunities fund the catch-up of the rest.

It’s also important to realize it’s not just an opportunity; it’s the preservation of history. Pretty soon, things I watched in the late ’70s and early ’80s won’t be watchable under traditional means. The Betamaxes and VHSes are going the way of the dodo very soon. There won’t be the physical ability to look at these things, so start with your high-value, high-touch, the ones that you know and just go from there.

Last week, I heard someone on stage at an event say that, “if you’re overwhelmed with all these tapes sitting in a facility somewhere, just get rid of them.” Do you agree with that? Should leagues or teams say, “You know what, there are some things that it’s just not worth the effort of digitizing, so let’s just let it go”? Is that too crazy?
It’s provocative, I will say that. Using the phone analogy, I know my photos are backed up to my iCloud or my OneDrive. I have ways to do it. So I would say I am not a fan of saying, “Throw it away.” Before you take that step, make sure you know everything that you at least have, because so much of the value is not necessarily in content known but in what you don’t know.

Content is being monetized in new ways; social-media platforms want different types of content. How do you see the explosion of monetization opportunities around archives and this content going forward? How is that landscape changing?
For us and our partners, it has been great. People’s thirst for content has not gone down, and how people are consuming content has radically changed. You don’t have two social-media apps; you have 10, and it’s both an opportunity and a challenge. The opportunity is monetization, but you have to meet your customers where your viewers are now. You have to be able to find, distribute, and track the content. It’s much harder to do now when you have to monitor multiple social platforms and streaming services. The opportunity is there, but it puts more onus on organizations to keep going forward and make sure their house is in order.

What do you see as the Veritone advantage in helping your customers engage with their customers to make life easier for everybody involved?
It all starts and ends with data. Everything that is going on in and around events is a different point of data. What I want to see may be different from what my next-door neighbor wants to see. We provide the data to understand what Sean likes to watch and when and ensure the content is accessible and searchable. You can build a workflow around it to meet those areas of personalization and needs. If you don’t have your data house in order and your content isn’t in that type of format, you’re going to be further behind.

We’re in this world of AI, and you guys are doing great things with it. Then there’s Generative AI, which is generating content. Does GenAI play a role within your organization yet? Where do you see it fitting into the Veritone world?
GenAI is an output of what someone would want to do and is going to be only as good as one’s data. We provide the building blocks: we take their audio and video and break them down into the necessary data. If someone wants to take that to make GenAI, to make net-new content, that’s a little bit more into the creative landscape.

Specifically, as it goes to sports, these are live events. Could you use GenAI to create programmatic highlight reels? Sure, no problem. Are you going to want someone to re-create the game? Probably not; you’ll want the real thing. We worked years ago on a commercial where they took the Williams sisters — them then and them now — and, based on how they were playing in different videos, broke it down and created a match of one versus the other. That was an incredibly clever application because it took real-life inputs and created something unique and creative without disrupting what actually took place or trying to fool people. You just don’t want to ever see it used nefariously.

Back to the original point of looking out the front window of the car, does that mean that a smaller organization, like a smaller college, can get started at a much lower price point, as opposed to saying, “We have to begin this by going backwards in time to 1935 and our first football game”?
Exactly, exactly. It’s the whole point. It’s much easier because, if you don’t do that first, you’re still creating the legacy problem. I say, start first this way, and then let’s go look over. It makes it a lot more palatable for anyone to come in there, and, frankly, you get utility right away versus what you had to do in the past.

What is the onboarding process for a college that says, “We want to do a deal with you guys to digitize our basketball games and football games going forward”?
It usually starts with, “Where’s your content?” Where are you filming? Where does it take place? Where can we get accessibility to the content? And then, usually, it’s about where the stream is going. Is it going to a server in your university? Is it going to a broadcast center somewhere? That’s usually where the conversation starts first. We’re cloud-agnostic, but we do most of our stuff within AWS and others. Everyone is slightly different, but, with any big or small institution, we can find a way to work.

Give us a real-world example of how you can make an internal social-media or production team more productive in terms of dealing with content.
A perfect example: a high school social-media team is filming a football game on a Friday night. That can be ingested directly into our core product, aiWARE. In advance, you can provide a set of rules: “Find me anytime No. 12 does something.” You can set up the rule, and the system can programmatically identify, clip, and distribute to Facebook, X, or wherever [it’s wanted]. What I suggest first is just to prepare them in a gallery, because I always like the idea of having a human in the loop to watch it, make sure it’s really what they want, and maybe add five more seconds to the clip. You can have that in near real-time, pull that, do a highlight reel, and even bring in a third-party data file so you can say, “Oh, well, this is No. 12; by the way, he threw for 278 yards, had two touchdowns, and ran for 42 yards and one touchdown.” Having all that data readily available is a great benefit.

2026 brings a Winter Olympics and the World Cup. It’s going to be a big year of sport. What do these even-number years mean for you guys in terms of customer engagement and content?
It provides a lot of opportunity. Especially looking at the Olympics and the World Cup, there are many collegiate athletes — we do a lot of work with the NCAA — so there are going to be a lot of commercial opportunities around it: different sponsors, different documentaries, and everything else. They tend to be great, busy years for us and for our partners. This is why, when they have a little bit more of their house in order, it enables us to capitalize around that. Especially for events like the Olympics, it’s not just the high-flying snowboarding event; there’s the luge. There are niche things that are covered in different areas here, providing unique opportunities.

You just mentioned customers’ having their house in order. Everybody likes to think their house is always in order, but it often isn’t. How does somebody know if their house is in order, to have an efficient, effective deployment and partnership with you guys?
I would say, if someone had a centralized repository where they know all of their content resides, they’re a step ahead. With most organizations, even many tech companies, if you ask, “Where’s all your data?” they might say, “Oh, it’s here.” But you don’t have any videos on someone’s hard drive? You don’t have any video that someone took on a device and didn’t upload it someplace? That’s usually where it starts. The main thing is, do you at least have control and know where all of your content is? If that’s the case, then you are a step ahead of everyone in having your house in order.