Welcome to the drill down. We've got the business stories behind stocks and move. I'm Cory Johnson, chief market strategist for Epistrophe Capital Research. Well, just ahead, big contracts, hyperscaler deals, a balance sheet reset, Veritone readies for a next chapter. We talk to the CEO. But first, sponsor time. The drill down brought to you by Era. Never miss another critical event or insight ever with ERA. Customize your company watch list and track key events, mentions and filings and more with an easy to use customizable interface. That's era aai erra.com.
All right, I'm Cory Johnson. Welcome to the drill down. We talk about the business stories buying stocks and the move AI models need data like chips need power and Veritone thinks they found a new supply. Ryan Steelberg is the CEO of Veritone, company behind data refinery and so much else. Ryan, glad to have you join us from Southern California.
Thanks, Corey. Weather's beautiful. Typical December in Southern California.
Yeah, Disneyland. We all know that weather so well. It's not the weather we get in San Francisco. It is. You've had a number of tech ventures. I want to get into Veritone, of course, but all based there. And any sane person would want to be in Orange County if they could.
It's more common to see a lot of tech companies now. Amazon has done a lot more investing in the area, but back when I started Adforce in 1994, right out of school, there were still orange trees around. Not that old. You kind of touched on it, it was a little bit of a gaming hub back in the day, but obviously a lot's changed. Thankfully, I've still found a way for us to build new businesses and remain in Southern California.
Well, it was down there. I don't remember from my research.
So, I moved back to Orange County after. But Adforce started after my sophomore year. We got funded. Quick fun story: I never knew what a VC was. We just started the business and my parents started getting a bunch of phone calls from different VC funds including Hummer Winblad and Menlo Ventures back in the day.
Kind of the rest is history. Ann Winblad and Mark.
Fiance of Bill Gates I believe at some point.
I know that they dated. I don't know they were engaged. Those stories go to the grave with those two. Let's talk about Veritone because Veritone has gone through some really big changes. I would challenge someone to read the 10K and walk away instantly knowing what the company's doing right now just because it's a lot of buzzwords. To prove this, I'll read from your 10K: your main business now is to address the ever growing challenges surrounding unstructured data. We developed AIware, our proprietary AI operating system, orchestrating AI models through a suite of powerful applications on a secure open platform to reveal valuable insights of vast amounts of structured and unstructured data. What does that mean in English? What do you guys do?
Well, let's start with the buzzwords. That statement is probably verbatim of what it was in 2014. And if you read our S1 in 2017, exact same. So what we do is there's lots of unstructured data being produced, primarily audio and video, which is the area of unstructured data that we focus on. We had a vision that there was going to be a need for a platform from ingestion to cognitive AI processing to indexing to create value or yield from all that unstructured data. You needed an end-to-end platform to turn messy audio and video into structured metadata. Meaning, if I'm analyzing 24/7 every video stream from ESPN and iHeart Media, how can I turn that into data, time-correlated data that machines can understand and act upon, whether it's for search and discovery or ad optimization or the speed at which SportsCenter is put together. That's what we set out to do. We initially were really focused on trying to optimize advertising, that's kind of my background, but from that beginning it expanded into many different types of use cases that Veritone supports today. The only other thing I would add is that Veritone means kind of a play on words, veritas and tone, truth in the signal, trying to get understanding from unstructured data. That's what we did. All those fancy terms you just read out of my 10K, in a nutshell, we're really, really good at ingesting hundreds of thousands of hours of audio and video every day and using AI to structure it, label it, and organize it.
And yet your business has changed a lot and you've gotten rid of a big ad business and a big pile of debt with it. Tell me what that piece of the business was. I want to focus on what you do now, of course, but I want to know what you jettisoned that was so big and why you did so.
It represented about 27% of our revenue. But it was an ad agency. Think of it as owning a pure services traditional media agency. It's still using our legacy namesake for a little bit longer called Veritone One. They merged with Oxford Road and they're the largest podcast media agency. I started that media agency as a separate entity, almost like a playground generating revenue where I could apply Veritone technology to this agency, which was a direct beneficiary of all this indexing. When we achieved enough scale on the core AI software side, the opportunity presented itself for us to sell that business for around $100 million, about $83 million in net cash, and we were able to jettison that business to help run operations and use a lot of those proceeds to pay off debt. What we did not sell was any of the software or the AIware business.
And you had a lot of debt with that which is now off your balance sheet. So people looking at historical financials might be confused by that big change.
We started this business in 2014 and it was early in the AI curve. Our main platform is called AIware and we never envisioned building N applications for our customers to use. We envisioned a world where the Disneys and the NBCs would give us all their audio and video, we would process it and index it and deliver it back to them with structured metadata via JSON. The market was early, so we started to move into the application business. Over that period of time, having an expensive engineering team, we needed a lot more capital.
What would you imagine back then in 2014 they would do with that data and how close to real time would they get it?
In near real time. But we were envisioning that their internal systems would be able to leverage the structured metadata to do a lot of things that ultimately the applications we built for them did, like optimize their ad placement, do research and discovery, for example what hosts are performing and where you're seeing burnout or turnover of their talent. We expected them, if we gave them the clean data sets, that their historical media asset management or media workflow systems would be able to use this data to get the value. That was a mistake. We learned that those legacy software applications were deficient. They couldn't just use our data; we had to go build a lot of that front-end application for them to derive the actual value from it. For example, iHeart Media has been a customer for many years. They have 12,000 users of our N applications. So we're not just delivering clean data back to their CTO and CIO. We created an application on top of AIware and we're servicing the end users as well. That's really where they started to realize the value of us indexing all this audio and video.
And that value would be realized by changing the content or matching the right advertiser with the right piece of content.
