Good afternoon. Welcome to Veritone's First Quarter 2017 Earnings Conference Call. Joining us for today's call are Veritone's Chairman and CEO, Chad Steelberg; and the company's Chief Financial Officer, Pete Collins. [Operator Instructions].

Please note that certain information discussed on the call today will include forward-looking statements about future events and Veritone's business strategy and future financial and operating performance, including its expected second quarter performance.

These forward-looking statements are only predictions and are subject to risks, uncertainties and assumptions that are difficult to predict and may cause the actual results to differ materially from those stated or implied by those statements.

Certain of these risks and assumptions are discussed in Veritone's SEC filings, including its registration statement on Form F-1 and its quarterly report on form 10-Q, which the company expects to file by June 23, 2017.

These forward-looking statements reflect management's beliefs, estimates and predictions as of the date of this live broadcast, June 20, 2017, and Veritone undertakes no obligation to revise or update any forward-looking statements to reflect events or circumstances after the date of this conference call..

Finally, I would like to remind everyone that this call will be recorded and made available for replay via link available in the Investor Relations section of the company's website at www.veritone.com. .

Now I would like to turn the call over to Veritone's Chairman and CEO, Chad Steelberg. Sir, please proceed. .

Welcome, everyone, and thank you for joining us today. After the market closed, we issued a press release announcing our results for the first quarter ended March 31, 2017, a copy of which is available in the Investor Relations section of our website. .

This is an incredibly exciting time for our company, both for reasons that are obvious and others that are less so. In terms of the obvious, as you're all well aware, Veritone became a publicly traded company on May 12. I think it almost goes without saying that we are thrilled and appreciate the investor support.

We're looking forward to the new opportunities this provides us by having access to greater depth of resources and a much larger and more diversified shareholder base. .

Now in terms of the less obvious. At Veritone, we talk a lot about our pursuit of building something bigger than ourselves, of making a dent in our universe. And when I say universe, many -- you may rightly think of the physical universe around us, the one directly in front of you that you can see and hear and touch, and that's true.

But at Veritone, we also operate in another important universe, the digital one. And something many of you may or may not know is that these 2 universes are merging into one.

As more and more devices and content providers bridge this digital divide, in the process generating zetabytes of new unstructured data, the human mind has become the limiting factor in terms of the scale, speed and breadth of understanding required. These digital universe of ours used to only speak one language, the language of computation.

But today, 80% of the data stored in our digital universe, including audio, video and image files that can't be understood with classic computation is only accessible to another kind of language, what we call cognition, the ability to perceive this data more like humans do, to recognize faces, voices, sentiments and objects, among much more. .

Cognition is something we humans take for granted, but to the machines that power our digital world, this was a code they couldn't crack until very recently.

The advent of artificial intelligence has brought us to a point where we can now empower these machines to mimic human levels of cognition, albeit at a more limited level for now, and to help us leverage the power of unstructured data in ways we never thought possible. .

It is our mission at Veritone to enable the machine that control our digital universe to use these new language, the language of cognition, the language of artificial intelligence, to understand both the digital and the physical universe unleashing a wave of innovation that, we believe, will transform our world. .

In a minute, I'm going to take some time to explain why we believe that Veritone is at the epicenter of this AI revolution inflection point. My hope is that by the end of this call, you have a much better understanding of what Veritone's business looks like and the incredible opportunity ahead of us. .

But before I do, I'd like to hand the call over to our CFO, Pete Collins, who will walk us through our financial results for the first quarter as well as the specific tie to our initial public offering.

Pete?.

Thanks, Chad, and good afternoon, everyone. It's a pleasure to have this opportunity to speak with all of you today. .

Turning to our financial results for the first quarter ended March 31, 2017. Our net revenues increased 50% to $3.1 million from $2.1 million in the same period last year.

The improvement in net revenues was primarily due to an increase in Media Agency revenues of $864,000 or 43% as well as an increase of $168,000 or 410% in SaaS licensing revenues from our Artificial Intelligence platform. .

Looking at our AI platform business. The total contract value of new bookings received during the quarter increased to $1.9 million compared with $156,000 in the same period last year. In addition, our monthly recurring revenue under agreements in effect at the end of the quarter increased to $111,000 in Q1 of 2017 compared to $14,000 in Q1 of 2016.

Although our revenue mix this quarter was more than 90% derived from our Media Agency business, which was consistent with our historical pattern, we plan to grow our AI platform over time, so that it will be the main driver of our net revenues in the future. .

Our gross profit for the first quarter of 2017 increased 66% to $2.9 million or 93.7% of revenue from $1.8 million or 84.6% of revenue in the first quarter of last year.

The increase in both gross profit and gross margin was primarily due to the operating leverage provided by our higher net revenue level as well as to lower transcription costs per hour, which were due in part to the higher volumes of data being processed. .

Our total operating expenses in the first quarter of 2017 were $9.5 million, an increase of 68% compared with $5.7 million in the same period last year.

The increase was primarily due to higher investments in software development and engineering as well as in sales and marketing as we continue to expand our AI platform business and enhance that platform, including developing new products and functionality. .

Our net loss attributable to common stockholders for the first quarter of 2017 was $6.9 million or $3.09 per share compared with $4.8 million or $3.86 per share in Q1 of last year. The EPS figures are based on 2.2 million shares outstanding for Q1 2017 versus 1.2 million shares outstanding in Q1 of 2016. .

Now turning to our balance sheet. At the end of the quarter, we had cash and cash equivalents of $8.8 million. A number of events related to our IPO positively impacted our balance sheet following the end of the quarter.

In addition to the $32.6 million in net proceeds we received from our IPO, which closed on May 17, upon the closing of the IPO, we received an additional $29.3 million from our largest outside investor, Acacia Research Corporation, which exercised its primary warrant in full to purchase an additional 2.15 million shares of our common stock. .

In addition, in connection with the IPO, the lenders under our recent $8 million line of credit exercised their option to fund the remaining amount available under the line, which provided us with an additional $4 million on top of the $2 million that was borrowed in March of 2017 and the $2 million that was borrowed in April of 2017.

Upon the closing of our IPO, all of the remaining principal and accrued interest under Acacia's $20 million secured promissory note and the $8 million line of credit were converted into an aggregate of approximately 2.1 million shares of Veritone common stock, and our preferred stock was also converted into common stock.

The total net proceeds to Veritone from the IPO warrant exercise and line of credit funding were approximately $67.9 million. On a pro forma basis, reflecting these events, at March 31, we had $76.7 million in cash and cash equivalents, no long-term debt and 13.9 million shares of common stock outstanding. .

That completes my financial summary. I'll now turn the call back over to Chad.

Chad?.

Thanks, Pete. To more fully appreciate what we're building and where we're headed, it's important to grasp the problems we are solving and the opportunity that it presents for Veritone. As I mentioned in my opening remarks, 80% of the world's data is unstructured.

This means that this throve of data cannot be easily searched, understood or utilized without human intercession. This problem is compounded by the fact that the volume of unstructured data being created is continuing to grow exponentially. .

The research firm IDC has predicted that the total amount of digital data created worldwide will double every 2 years to 44 zettabytes in 2020. That is an almost inconceivably large amount of unstructured data. A zettabyte is 1 trillion gigabytes.

And it's going to require this new language of cognition powered by the artificial intelligence to unlock its full potential. .

The problem is that global organizations, both big and small, lack the proper tools to analyze the growing variety, velocity and volume of data. This is where AI systems, particularly cognitive systems based on machine learning, come into play.

These systems are being designed to make sense of the complexity and unpredictability of unstructured data at volumes and speeds that exceed the capabilities of the human brain. The issue with these systems, however, is that they are only useful when applied to a single narrowly defined problem.

For example, you can teach a computer to recognize faces nearly as well as a human can, but that same computer is not capable of recognizing objects or spoken words. .

When my brother Ryan and I started Veritone in 2014, our founding mission was to bring the power of AI-based computing to companies of all sizes by providing a unified cloud-based platform to satisfy a wide variety of use cases across many industries.

We began investigating and integrating as many of these cognitive engines we could find, which, at that time, was only a handful. And we're amazed by the power they possessed, but disappointed by the narrow cognitive capabilities of each engine. .

As a result of these early findings, we hypothesized that by integrating multiple discrete cognitive engines into a common framework, could we create an ecosystem that would have the depth and breadth of intelligence to solve much more complex and valuable problems.

As a result, we expanded our engineering efforts, which has led to several groundbreaking discoveries. One of these was a new architecture for storing short- and long-term cognitive data, a sort of synthetic memory layer.

And another was an intelligent router that can orchestrate multiple cognitive engines and govern the cognitive engine selection process. .

The potential benefits of this are multiplied greatly as the number of available cognitive engine increases. While we could only find a handful of engines in the market in 2014, at the end of March 2017, we had 48 active engines on the platform.

And we are tracking over 5,000 other engines across 40 cognitive classes, and that number is scaling at an incredible rate.

As the number of cognitive engine in each classes increased, we have seen incredibly rapid improvement in their speed, accuracy and cost, with existing engines improving at different rates and new leaders emerging almost continuously. .

To manage all these disparate cognitive engines, you need an AI operating system that continuously learns and leverages the unique performance characteristics of each, while also optimizing the cost-benefit yield curve for our customers.

We theorized and have now shown that by intelligently running multiple engines in the same cognitive class, we could exceed the accuracy of any single engine. .

The Veritone Artificial Intelligence Platform, or AIOs as we like to call it, takes an entirely new approach to AI by combining the capabilities of multiple third party engines, each designed to perform a specific task.

The Veritone Platform makes significant strides towards the capabilities of so-called artificial general intelligence, which is the still unrealized goal of 60 years of AI research to create a machine that can perform all cognitive tasks as well as the average human. .

Our ecosystem of cognitive engines can be orchestrated together and continuously evaluated to provide the best possible outcome depending on the use case. We expect a number of engines on our platform, which totaled 48 at the end of the quarter, to continue to grow at an accelerating rate. .

It's important to note, however, that from time to time, we elect to drop engines that we deem are no longer up to our rigorous standards. And in order to improve the performance, cost and speed of the data analysis process, we are developing proprietary conductors that orchestrate the entire ecosystem with 0 human interaction.

This is machine learning applied to machine learning. We have launched our conductor for transcription engines and are working on additional conductors for other classes, such as object detection. .

Veritone's AI operating system comes preconfigured and bundled with a native application suite to complete tasks, like content management, discovery, sharing, administration and analytics. .

As with Salesforce, no platform is complete without third party apps to make it useful and extensible to the end user. Veritone is working with our partners to develop applications to expand the power of cognition to new markets. .

There is obviously an enormous opportunity in front of us. To put some market research around the size of the more immediate opportunity, IDC predicts that the worldwide revenue from cognitive systems and AI will grow from $8 billion in 2016 to more than $47 billion in 2020.

Obviously, the market for cognition systems represents a subset of this total, but is a large and fast-growing market. .

As some of you know, Ryan and I have built several successful digital advertising businesses. So when we initially developed our AI platform, we first deployed it in our Media Agency services business, Veritone One.

Today, that platform helps our advertising clients, including Uber, 1-800-Flowers and Dollar Shave Club to maximize their ROI on their ad spend using realtime ad verification and media analytics. .

As Pete mentioned earlier, it's important to note that since our inception, we have generated most of our revenues from our media services, but we expect that mix to shift over time to where our pure SaaS revenue from our AI platform will eventually provide the majority of our net revenues. .

first, the number of new customers added under master services agreements; second, the total number of active customers; and third, the average media spend per customer. .

Our first 2 KPIs in the business, which track the number of new and active customers, provide us with insight regarding our ability to grow the market share of our Media Agency business by winning new customers as well as keeping tabs on customer churn.

During Q1, we added 8 new net customers compared with 8 -- sorry, compared with 4 net new customers in the same period last year, a 100% increase. We had a total of 39 active customers as of the end of the quarter compared with 30 at the prior year-end, a 30% increase. .

The third KPI, average media spend per customer, allows us to analyze not only spending trends, but our ability to grow media spend with existing customers. During Q1, average media spend per customer was $670,000 compared with $602,000 in the same period last year, an 11% increase.

It's important to keep in mind that while this business is mature and provide a solid foundation for Veritone, it can also experience volatility in net revenues from time to time. .

After realizing the power of the platform in the media market, in early 2015, we expanded our focus to other verticals, politics, legal and law enforcement and public safety, where we believe that cognitive computing has the power to transform the way companies, organizations and public agencies use and manage data, the same exact platform and application serving disparate markets.

And this is just the beginning of the industries that we intend to transform. Through our recent announced partnership with Quantum, we've now also entered the data storage and management space.

Our groundbreaking Veritone AI ware, a hybrid on-premise and cloud version of our AI platform that is scheduled for release in Q3, will be offered as integrated solution with Quantum's StorNext workflow storage.

This combination will allow users to leverage the power of our cognitive platform to extract new value from their on-premise video and audio content without having to store it to the cloud. We are extremely excited about this partnership and look forward to reporting to you on its success in the months ahead. .

The business model of our AI platform is centered around bringing the power of cognitive computing to a broad range of industries. What I love about Veritone is that our revenue and, ultimately, long-term business value is directly correlated with the value that our customers derive from our platform.

Specifically, the amount of content our customers process, the performance and breadth of our cognitive engine and the variety of applications servicing our end users are the primary drivers of customer value. .

number one, the total number of video and audio files ingested and processed on our platform; number two, the number of third party active cognitive engines on the platform; and number three, the number of accounts on our platform. It's important to keep in mind that while these metrics will grow over time, it will likely grow at different rates.

And their growth may not always be linear. .

Tracking the total hours of video and audio files ingested and processed on our platform, which was 367,000 hours in Q1 of 2017 versus 190,000 hours in the same period last year, an increase of 93%. The increase in the total hours provides insights into our customers usage of the platform. .

The second metric is the number of third party active cognitive engines on our platform, which was 48 at the end of the first quarter of 2017 versus 11 at the same point last year, an increase of 336%. This is an important metric because as I mentioned earlier, it indicates the depth and breadth of cognition on our platform. .

And finally, the third metric is the number of accounts on the platform, which increased significantly to 57 as of the end of the first quarter of 2017 versus 4 at the same period last year.

The increase does not include our group-wide deal with CBS Radio, which licensed our platform to provide its 117 radio stations with near realtime ad and content tracking, comprehensive analytics, faster content extension and smarter media management.

This group-wide deal added 110 accounts in the second quarter of 2017, above the 7 CBS Radio accounts that were active as of the end of Q1 2017. .

As we grow and expand our AI platform, we think it's important to provide goalpost to track our success. We expect to end the year with approximately 425 accounts and approximately 80 active third party cognitive engines on the platform, and having ingested and processed approximately 3 million hours of video and audio files during the year.

We expect that the integration of our platform with kCura's Relativity platform will significantly increase the number of accounts and the number of hours of video and audio files ingested and processed.

The integration with Relativity and the ramp-up of the sales process in this market have taken longer than we anticipated, which will impact our second quarter results.

However, we remain confident in our target for those 2 metrics for the full year because our platform provides a solution that has greater cognition at a lower price and other services available to that market. .

Obviously, with something like artificial intelligence, it's very difficult to talk about it lightly, so I appreciate you all for taking the time to get an overview of how our business and technology works. It's certainly a lot to take in.

If you're interested in learning more, we have a lot of great information on our website regarding our platform, our target vertical markets and some of the many use cases for our platform. .

At Veritone, we're looking at the problems of today and the future. We're coming up with real and innovative answers to solve them.

As our expanding customer base processes more content using a growing number of cognitive engines running on our platform, we expect that we will generate greater revenue and create higher long-term value for our stockholders. At our core, we remain committed to unlocking the total power of the digital universe to help businesses and people thrive.

Again, thank you all for your time today. We look forward to updating you on our progress on the future calls. .

We're now ready to open the line for questions.

Operator?.

[Operator Instructions] Our first question comes from the line of Bill Choi with Wunderlich. .

Want to follow up on that last comment about kCura. I know that for us, legal vertical represents a huge opportunity, particularly for big lawsuits and tracking tons of hours of audio tapes and so forth. So just wanted to get a little more color on that. I think as of sometime in Q2, you signed up 18 providers.

Relativity total partner base is about 120 with a lot of room to grow here. Just step us through what's been happening as you enable the providers and how some of the deals might look in second half as this tends to be a little bit more transactional, so any visibility on that legal vertical will be helpful. .

So Bill, it's Pete. So we're -- as Chad said in the prepared remarks that the ramp-up, both from a technical perspective as well as a sales perspective, has taken longer than we expected. The reseller agreement that you quoted are in place, and that's been accomplished.

And the -- I kind of -- I talked sort of the bugs as it relates to the technical integration has also been completed. So we're -- from our perspective, the ramp-up is, like Chad said in the remarks, a little longer than we had expected, but we're getting there.

And the good thing is that as we become more and more involved in that market with our General Manager, Michael McDonald, he's confirmed the hypothesis we had coming into the year, which is that we provide a very cost-effective solution and a solution that has much broader cognition or much broader depth of information for the users in that market.

So that's why we made the comment in our remarks that this longer ramp-up does not impact our overall goals for the year. We're still confident that we can achieve our goals for the year, despite the slightly longer ramp-up that we have experienced. .

Beside the technical aspect of it, is there anything that you need to do to get customer acceptance and usage up? And if there are any hurdles that you could discuss. .

No. This is Chad. We don't see any hurdles in front of us. In fact, just to kind of go back on Pete's remarks on the technology front, some of the feature sets that took a little bit longer to integrate, specifically, were the features of redaction, which is the key feature for the legal industry that we put into the platform.

It was -- the innovation that we built around that, actually, we filed a patent there as well given what we'll be able to do with our cognitive platform. So we see no reason on a business standpoint that we should have any further delays. It was all technically related. And we expect to hit our end-year objectives. .

Okay. The other question I had was the press release that came out a day or 2 ago announcing the new division. There was an update that you have 65 cognitive engines now enabled, so certainly quite significant growth just this quarter alone. And the bigger question here is how you go about monetizing this huge base of cognitive engines.

It's primarily been transcription that's been driving much of the revenue. How do you see that evolving as that cognitive engine ecosystems just seems to grow quite significantly? If you could tie that with some kind of customer adoption and any kind of contract or revenue, that will be helpful. .

Yes, thanks, We actively track the number of cognitive engines in the market.

Our current count exceeds over 5,500 third party cognitive engines and our goal at Veritone for the developer hub is to allow third party cognitive engine developers to join our ecosystem by a self-service model, providing them with a faster path to monetization, an access to our training data to enhance the performance of their engines.

We are using in-house marketing resources to spread the word in the community via tools, that is white papers, newsletters and press releases, to continue to expand that market. With the growth that you've seen last -- in this current quarter, we expect that growth to continue.

And our monetization strategy around that is really based upon our customers' demands. When we first started the business, it was really Ryan and I and our engineering team randomly selecting engines that we thought were interesting and could add value to our prospective customers.

And now as our customer base is diversifying, it really becomes a pull function, where they're reaching and communicating with us, specifically our Head of our ecosystem development team, Tyler Schulze, to reach into that pool of 5,500 and start onboarding specific engines that can add value to the problems in their market. .

In terms of -- what's maybe 2 or 3 major cognitive engine classes that could generate revenue for you besides subscriptions, if you have any thoughts on that?.

Yes. We are increasing customer demand for object detection, transcription analysis, facial recognition and detection engines.

The specific demands are resulting in highly targeted engines within these macro categories, which continue to reinforce the power of the Veritone Platform with speed to onboard new cognitive engines and integrates the resulting intelligence with other pre-existing engines as powerful and our key differentiator. .

Okay. One last question for me, then I'll pass along. One of the thesis for your capital raise was the ability to do accretive acquisitions and speed along customer adoption, require guys who might have customer relationship in some data, but then to supercharge it with your cognitive engine base.

Can you talk about the environment for potential targets and how the environment currently looks?.

So Bill, we're looking for businesses that process audio, video or other unstructured data and provide their products in a software as a service model to their customers. Those are the ones that are most interesting to us at the present. Our focus is limited in the verticals that we currently serve.

So companies that have significant existing customer bases are also very interesting to us, particularly where we believe that our platforms, technology and capabilities would enable those businesses to enhance the value of the solution they offer to their customers. .

And our next question comes from the line of Mike Latimore with Northland Capital. .

On the CBS Radio win and deployment, I guess, is that going to be at a full revenue run rate, say, in the third quarter?.

Yes. We expect to complete the rollout to all 117 stations by the end of this quarter. .

Okay. And then I noticed this in the press release. You talked about a win with a financial news, TV network.

I guess, can you just elaborate a little bit on that? Is the -- what kind of function are you providing there? Is there opportunity to upsell over time?.

Yes. We can't share the name of that customer because of the confidentiality clause in our agreement with them. But the customer is using our platform to analyze their broadcasts and provide their advertisers with the complete understanding of the benefits that those advertisers receive by being on their network. .

And then I think same with yesterday, you talked about Veritone Government. maybe, I guess, a new division. Can you talk just a little bit more about that? We've probably seen them some work in politics and public safety, but maybe if you can talk a little bit more about what opportunities you're seeing here.

It seems like it's a little bit more safe side of the government focus. .

Yes, absolutely. At the present time, we are managing the public safety and government sector for both local, state and federal services under a single business unit with subdivisions with dedicated leads targeting the discrete submarkets.

As you know, the problems facing public safety today are no longer fall neatly within classic geographic or even departmental boundaries. .

Okay.

And then just in terms of operating expense, should we assume operating expense in the second quarter similar to the first quarter?.

Yes. We're -- Mike, we're -- the guidance that we provided is really at the level of the net revenues and the drivers of that, that Chad mentioned, number of accounts, the number of cognitive engines and the amount of media that will be ingested and processed for the year. So that's really what we're focused on from a guidance perspective. .

Yes, okay. And I guess, just the last one on the bookings in the quarter. I guess, looks like a good number there.

I mean, is that sort of more bookings than you've had all of last year?.

Yes, it was. That was -- that would be the high watermark. Last year, our total revenues were in the agency -- I'm sorry, in the AI platform business were under $1 million. And to do total contract value of $1.9 million in the quarter, that's the largest amount we've ever done in our history. .

And our next question comes from the line of Joe Fagden with Craig-Hallum. .

Another one on the government business.

Just in general, are the contracts you're going to be ratable, like you have with like the media and whatever, is this going to be more of kind of a project-based type of vertical similar to the legal?.

first, body cams and dash cams; two, some of the municipality cameras and feeds that are coming in; and then third is the public media index from -- that's happening out there from both consumer content that comes in the law-enforcement agencies. So we see that as being a very constant staple consistent with sort of our media business.

But at the same time, as cases arise and large events happen, we'll see much more project-based work, probably, also impact the system. So it's a bit of a hybrid with the baseline of consistent content coming from their day-to-day operations. but also spikes involving specific events. .

Okay. And then I guess, as much color as you can give, over the near-term call, like through the rest of 2017, I think media and, perhaps, legal are kind of expected to be the 2 big verticals for you.

Do you think government can get, like, on that level? Or is it going to be a much more minor part? I guess, I'm trying to figure out how much this really expands your opportunity over the next few quarters?.

Yes. This year, it's a build year for us in the public safety sector, both for federal and local as you go through the various compliance and regulatory bodies to onboard customers there.

At the same time, we continue to harden our infrastructure to ensure that it's complying and residing in [ hardened ] secured data centers that meet both federal and local law enforcement requirement. So this is a build year for us, but we expect next year to be one of our primary new growth drivers for both revenue and expansion opportunity. .

Okay. And last one, I think I already know the answer to it.

The margin profile, the economics and pricing in the government vertical, is it the same as your other verticals? Are there anything else different with that?.

We believe it will come in line with the other verticals that we currently operate. .

thank you. At this time, this concludes our question-and-answer session. If your question was not taken, you may contact Veritone's Investor Relations team at veri@liolios.com. I'd now like to turn the call back over to Mr. Steelberg for his closing remarks. .

Thank you for joining us on today's call. As I mentioned, we are excited to now be a public company and want to thank our employees, partners and investors for supporting us through this transition and allowing us to scale our company even further. We look forward to updating you on our progress on our next call. Thank you.

Operator?.

Thank you for joining us today for Veritone's First Quarter 2017 Earnings Call. You may now disconnect..