Veritone, Inc. (VERI) Earnings Call Transcript & Summary

All right. Welcome back, everybody. So look, we're just delighted to have Veritone joining us here at the JMP Tech Conference. On screen with me is Brian Alger. Brian runs the Investor Relations and Strategic Corporate Development for Veritone. Brian, welcome.

Thank you. It's good to see you.

How is your day going?

It's going well, a little hectic.

I bet. I bet. And where are you today? Looks like...

Yes, I'm physically in Northern California. Can't really complain up here in Wine Country.

Yes. Yes. That's awesome. All right. So tell us a little bit about yourself. Where are you from? What did you do before Veritone?

Oh, God. I'll save you all the gory details, but immediately...

Where were you born? Tell us that part, where were you born?

Born in Fukuoka, Japan. My dad was in the Military.

That's a lot of conversation.

Right. Yes. So a little known facts about Brian Alger, yes. But I spent most of my career on The Street doing what you do as a sell-side analyst and spent about 10 years as a hedge fund manager. Prior to joining Veritone, I worked with a team over at ROTH Capital, both as an analyst and a banker. And through that engagement, I got to know the team at Veritone, both pre-IPO and then post-IPO. And over the course of a number of years, Chad recruited me into the company and just under 2 years...

Was it a couple of year recruiting process?

We danced a little bit. It's kind of hard to leave The Street when you've been stuck in it for so long. But I joined about 2 years ago and made the jump and haven't looked back at all. It's been a great ride.

All right. So talk about what the opportunity looked like when you joined 2 years ago? What the positive surprises were and negative surprises were when you joined and then we'll sort of unravel the 2 years.

Yes. It's -- clearly, I was betting on a horse, I was betting on an opportunity to make a jump. And in reality, the story from Veritone pre-IPO hadn't changed at all. There've been -- growth rates that didn't get met necessarily right out of the gate. But the story had been the same, the strategy had been the same and I had been a believer from day 1. And basically, about 2 years ago, I started seeing the pieces come together with the state and local markets starting to get some specific products delivered to them, specifically IDentify and Redact were coming to market, and there was early signs of Automate Studio that really got me excited about the potential. So I made the jump a couple of years ago. And we've been executing against the same strategy...

So this is 2018?

Yes, '19, beginning of '19.

2019?

Yes, beginning of '19. And it's been, in many ways, a lot like I expected. In other ways, maybe not so much. But it's -- when you go from working on Wall Street and you're focused on investing only and you're always looking at financials and things on a quarterly basis, you don't always see the inner workings of the sausage making that goes on. And I knew that, that would be a change and it was. And I could say now that I've been in it for a little bit, I couldn't be more happy to be here.

Awesome. Okay. So -- I mean let's talk about what -- sorry, the big -- the business, in the last 2 years, has dramatically cut the burn rate, has seen the growth really accelerate. Let's talk about what you think, from your perspective, the sort of 2 or 3 biggest drivers in that improvement in the financial fundamentals has been?

I think there's a number of things. At the end of 2019, the company made an important shift. It changed its way of looking at the business. And at the end of the third quarter, going into the fourth quarter of 2019, there was a realignment. And that realignment adjusted the cost structure a little bit, and that certainly helped on the burn rate. But more specifically, and what The Street wasn't aware of at the time, was that we had made a big shift technologically. We were in the midst of deploying V3 of our operating system, our third-generation of the operating system. And what that enabled was a massive improvement...

And this was in '19?

It was at the end of '19 and implemented at the beginning of '20. And that V3 implementation across both of our cloud providers, AWS and Azure. But more importantly, now available at the edge on-prem and virtually anywhere. We would like to say, aiWARE Everywhere, with our relationship with NVIDIA, which you and I can get into. But all of that was enabled with V3. And with V3 coming into the market at the beginning of 2020 not only were we able to get real traction and real growth with our government and legal customers, both at the state and federal level, but we also were able to expand it in new markets. And I think it's that new market penetration, specifically energy, clean tech energy, even more specifically, that woke The Street up to the reality that Veritone is a platform play. It is an operating system company, first and foremost. And end markets that we happen to monetize just happened to be the markets that we went into first, are SAM, if you will. The markets that we can serve, we believe, is bigger than anybody else on The Street because of the operating system. We're not tied to services. We're not tied to channel partners or anything else. It's a technology-first play that has enormous scale and extensibility.

All right. We're going to get into that more. That was part 1. So part 2 is -- and I'm not going to ask the how's business question. For anyone who doesn't know, Veritone's currently in their quiet period, so we're going to avoid putting Brian in that situation. But I am going to ask the other 2 that I've asked all the other executives over the last 2 days, which is -- number one is just if there's one thing that you guys did in 2020 that you had not done before but that worked really well, what would it be? And just for context, some of the other answers we got, we got one CEO, who said, I used to have the Town Hall meeting every 2 months. COVID hit, I started doing it virtually every week, and I started out sending video messages, 2 to 3 minutes max, every day, every other day, and he goes, they worked. Now -- he said in hindsight, I was too scripted on the ones that were every 2 months and it was less optic. Because now I do it every 2 weeks, but that was -- and then we had another company say, we used to be really event-driven and so we get people to come -- thousands of people to come to our events, and then that would be our lead gen. And when that went away, we started to -- we actually built a customer database, an ideal customer profile and we started [indiscernible] which products, which customers and we're running a lot better as a result. So those are just a couple of examples.

Yes. Yes. Well, clearly, 2020 was unprecedented on so many levels. For us, we went work from home corporate-wide globally, actually before California even went into lockdown, in the first week of March a year ago. So almost a year to the day. And there was, at that time, a lot of uncertainty about how our employees would react to the situation, but also how everybody else was going to react because we...

You guys went 100% work from home. No one's been in the office basically for 300 days?

Correct. Correct. And that's still the case. And we had good confidence in our ability to communicate and to maintain employee engagement and employee satisfaction, but we really didn't know how society in general was going to embrace the change. And we've talked about this at the executive level quite a bit as we're kind of all reevaluating ourselves and I think the basis for your question is how do we move forward. We're looking at it. And it's astounding that society was ready on a whole to see so much transformation to do things like this, Zoom call, a virtual conference. We all used to travel all around and have to hobnob in various hotels around the country. But we found a way, right? And at Veritone, we did more than that. We managed to actually improve our productivity, we managed to improve our employee satisfaction, we managed to obviously grow our business and engage with our customers more. So that, on a whole, it's just that society was so adaptable. I think it's probably -- if I have to succinctly say it, the society as a whole was able to adapt to such a radical shift.

All right, awesome. So it sounds like you might stick with it. That's what I'm hearing.

We're evaluating that. I don't think anything definitively has been decided at this point. For those that pay attention to SEC filings, they'll note that we are subleasing out our headquarters. As of last night, we filed an 8-K to that regard. So we are evaluating how much space we need and what that space is going to be used for.

The one that's near -- that's in Orange County?

Correct. The Costa Mesa headquarters has been subleased out. And candidly, we've had a couple of other facilities. We have offices in New York, Denver, Seattle and Costa Mesa. And obviously, I'm up here in Northern California. But we're evaluating what we need. There's going to need to be hubs because we do need to get together with our customers. We do occasionally need to get together and collaborate with our design teams. But it's not going to be the same. It's going to be some combination of what we've realized in terms of new efficiencies and productivity and also some face-to-face and in-person capabilities.

Yes. I'm going through the filing now while you're talking, but maybe just tell us what it says.

I actually don't know what it says, to be candid. I know that we filed it. I know the particulars behind it. But the net-net of it is we have subleased out our headquarter space and we'll be evaluating what we need for -- down the road.

Okay. Well, maybe some of it. If one of my associates is on this call, this might be a good time to look that up, and then he can ask a question about it in the next...

Well, for instance, I don't know if the 8-K says who we subleased it to. So I'm just going to hold off on that.

All right. We will find out. Okay. Cool. That moves us to the part -- you basically answered the 2 questions at the same time for part 2. So part 3, current events. And again, we're avoiding actually asking you how your quarter went. So I think the big question that a lot of investors have is just there's you guys, there's Palantir, there's C3.ai. And then there's a host of -- I was going to say smaller, but actually, some of them are not small at all. There's a host of private companies that are probably going to be coming to the market. So let's start with Veritone versus Palantir versus C3. How do you think investors should differentiate these companies?

Well, in many ways, we're very, very different. I mean obviously, we're all using cognition, machine learning, AI, more broadly, to enhance and augment business processes and digital transformation, right? So in that way, we're similar. But the go-to-market is -- couldn't be more different, right? Just to think about it. Palantir has been around since what 2003, right? C3, I think, was started in 2009. Veritone was started in 2014. By definition, we're cloud-native. I think the others have evolved their businesses. Because we started in 2014, we had the benefit of starting with math breakthroughs that were made in 2010 and 2011 and '12, both from a software and mathematic standpoint as well as hardware architecture, particularly in deep learning and neural networks. So from the get-go, when we formed our operating system, we had perhaps a better starting point or a more well-defined starting point. But really, when you look at the go-to-market, it's 3 different proven business approaches and time will tell which one is going to be the most successful. But the first was Palantir, coming out with largely a services-driven approach to very, very, very complex problems with very large organizations and applying a lot of human capital to figure out how to leverage mathematics to solve those problems. And they've been obviously successful in doing that. C3 has taken a very traditional enterprise software approach to it, where they've taken some elements of technology, applied channel partners and traditional enterprise sales approaches to it to...

Sorry, say that again. What was the beginning part? How they -- what did you say the C3 approach was?

It's basically leveraging pieces of -- it's traditional enterprise software sales, right? It is -- it's tried and true, where they're leveraging pieces and elements of technology with channel partners and their own go-to-market sales team. And they're finding big complex problems, like working on maintenance or predictive maintenance issues and leveraging partners for that. The approach that we've taken, and we think it's actually very well-proven, is an open standards-based operating system to build a platform to enable rapid ecosystem development. So allowing for the rapid development of AI models, we call them engines, across -- right now, 28 different classes of cognition that we currently support, allowing that vibrant ecosystem. You talked about the private companies that are out there. Many of them are developing very, very good, innovative machine learning algorithms that are very good at very specific niche applications. Applying them and leveraging them with our operating system, they can be instantaneously plugged into new application development because we have a standard API layer, like any other true operating system. And through that API layer, the software developers, ourselves included, as well as third parties can develop the programs agnostic of what engine or what model is being utilized underneath for the classic cognition. You just simply say, I want to do a translation or I want to do a predictive model or I want to do object recognition. And the operating system will, through the standards approach, identify what model is appropriate for the necessary language. And we do that in a low-code, no-code tool that we call Automate Studio, which really accelerates the product development. And that approach, candidly, is something that we've seen time and time and time again. I mean you go back and you look at mainframe computers versus Windows for the PC, right? In the early days, there was a very large centralized solution to very complex problems that ultimately, through democratization, resulted in a very, very big opportunity. Same thing occurred at the data center, Linux versus, you could say, Solaris at one point in time. And more recently, Android versus iOS. Android is massively more pervasive globally. And the ecosystems around all of those operating systems have enabled very, very rapid development and extensibility into virtually every segment of society. So today, we're a relatively young company, certainly versus those 2. But we're serving a multitude of end markets in a multitude of different ways, and it's the operating system that enables that. And that approach and that go-to-market is what really sets us apart from, we think, anybody else in pursuing the AI market.

All right. Super helpful. All right. We've got 6 minutes to talk about 3 topics. So we got here NVIDIA, Energy/clean tech and the partnership with Alteryx. I'll let you pick which one we go first.

Let's do all 3, and I'll do them quick. So all of those have been announced. All those are public. So we'll start with Alteryx because I think it's just clean and simple. Alteryx is a company you cover, and you know it really, really well. And I think...

I'm actually speaking at the exact same time right now.

Well, there you go. We use Alteryx. We are customers of Alteryx. We think the world of them. They're very, very good at what they do. However, as some people recently have pointed out, there are some things that they don't have in their bag of tricks yet. We bring that to the market. So we bring, through Automate Studio, low-code, no-code programming capabilities. Through our extensibility into the cloud, we bring infinite scale. And we also bring cognition of unstructured data in the exact same desktop environment that Alteryx users are accustomed to. So all 7,000 customers currently using Alteryx all around the globe, albeit every shape and size, now have new capabilities unleashed by AI through aiWARE and what we bring to that integration. We're very excited about it.

I'm going to elongate this, already excited. For example, you can pull an audio file into Alteryx, right, and have it processed in the cloud through Veritone and then get the results back in Alteryx?

Yes. And it's not just audio and video. It's any type of data. We can apply cognition against it. And Alteryx is just the first, right? We're going to do this with many others. There's lots of people in the data analytics world that we see our ability to assist or augment what they already do. In the case of NVIDIA, I'll jump to that. NVIDIA, I think everybody would recognize, is the foremost leader in hardware -- AI-specific hardware. And Jensen, I've known really, really well since before he went public. He's been pursuing this forever and he's built an enormous ecosystem of partners on the software development side, on the mathematics side, starting initially with high-performance supercomputers. And now through his new platforms, he's extending into robotics, automation and now to the edge with his EGX platform. We went through a period of time where we had to figure out how to take the cloud-native containerized AI capabilities that we bring to the market and figure out how to then leverage CUDA, which is the kernel, if you will, that NVIDIA utilizes to standardize across all of its platforms. We've done that now. And in effect, we now have enabled anyone to develop against our operating system down to the CUDA-enabled NVIDIA graphics card systems, whether that's for a supercomputer, whether that's for an edge computer like EGX or if it's even down to the Jetson modules in their AGX platforms. You can have that same code base because of the standard API layer to develop against. And really what you get there is a Wintel like type of development environment. So for us, we're able to go anywhere that NVIDIA sees the market going, and that could be for IoT, that could be for energy, it could be for health care, it could be for any number of high-performance supercomputing applications, big data analytics, et cetera. The key is it's a standard API layer to develop software against a unified layer underneath with CUDA-based GPUs. So as that applies to energy, when we're taking our...

Did you cover NVIDIA by the way when you were on the sell-side?

I did, and I did it well. Yes. So I do know the architecture.

You seemed like it. Yes.

Probably more than most.

More than I do, that's for sure. Okay. We got 2 minutes to the last one.

So in energy, I saved the best for the last. Look, energy is huge. Clean energy is even bigger. You look at what the Biden administration is talking about, about revolutionizing the energy space. In order to do that, you have to optimize the grid and you have to recognize that we have existing infrastructure that has to adapt to the new markets. In order to do that, you have to do it at a sub-second scale and you have to do it on a distributed basis. The only mathematics that can possibly work for this is a Hamiltonian-based model. We happen to have the leading expert in Hamiltonian-based models for predictive analysis and control. That's what we have. It's based into our 5 different products. I encourage all of you to go to the website and read up on it. But in effect, what we're delivering is sub-second control of forecasting and prediction to be applied against any size of grid. It could be within a single home, it could be for a microgrid or it could be an entire regional grid, like the one that we're currently deployed with right now.

All right. That's awesome. And I mean this space is -- there's so much going on in this space right now, right?

We're down to 25 minutes.

Can you give us a concrete example of what a use case would be?

For energy?

Yes.

Sure. So right now, we leverage IR sensors around the solar array to, in real time, predict how much energy is going to be captured by that solar array. So if a cloud is passing overhead or if maybe it's a foggy day because there's more humidity in the air, we get real-time analytics, real-time prediction in terms of the production capability and we pull in the usage data in real-time off of the smart meters. And through that, we can basically optimize the flow, not only through the grid but also how much needs to go into spinning reserves, think of that as storage, or smoothing capabilities. And our controller software allows for that to be done on a sub-second basis. So we're doing that right now in real-time with our lead customer.

Wow, that's awesome. All right. Brian, thanks so much for joining us, and we really appreciate it. And we're looking forward to seeing the results.

Look forward to it. Thanks, Pat.

All right. Bye-bye.