C3.ai, Inc. (AI) Earnings Call Transcript & Summary

Good afternoon, everyone. Before we get started, if you are a member of the press or media, please disconnect at this time. This is a restricted line. Any unauthorized party in this meeting or any unauthorized use of the information communicated in this meeting is subject to prosecution to the fullest extent of the law.

Hey, there. Good morning. Thanks, everyone for joining us again for day 3 of the Wells Fargo Technology Summit. I'm Michael Turrin, software analyst here at Wells Fargo. Joining us for next session this morning, we're very pleased to have a name well-known in software, Tom Siebel, Chairman and CEO of C3.ai. I'll start off with some questions. If anyone has anything they'd like for me to ask or loop in into the conversation, just shoot me an e-mail, I'll do my best to add those into the conversation as well. But before we get started, Tom, thank you very much for joining. I appreciate your time.

Thanks, Michael. Nice to see you.

Likewise. So I think it's very interesting to start with the overview. This is clearly an idea that you've had for some time. So can you talk a little bit around the vision C3 was founded upon? And just provide an overview for those investors who might not have the complete picture around what it is you do and some of the customers you serve today.

Thank you. Well, C3.ai is a computer enterprise -- computer application software company. And the big idea was to build a software platform that would enable companies to take advantage of elastic cloud computing, big data, the Internet of Things and predictive analytics, okay, in a way to build large-scale enterprise applications. So we started in 2009 and spent the better part of a decade and about $1 billion building this software stack that we call the C3 AI Suite that provides all of the services necessary and sufficient for organizations like the Missile Defense Agency or the United States Air Force or the Army Futures Command or Royal Dutch Shell, Koch Industries, Bank of America, what have you, to design, develop, provision and operate at very large-scale enterprise AI applications. On top of that, we've built a family of about 40 turnkey applications for the financial services industry, for the telecommunications industry, manufacturing sector, oil and gas, utilities that address the entire value chain of those industries. For example, we have turnkey applications in financial services that handle problems like cash management, credit approval, risk scoring, anti-money laundering, fraud detection, Volcker Rule compliance. We have a family of applications for grid operators -- power grid operators that basically address the entire value chain from power generation, transmission, distribution, integration of renewables, energy efficiency, what have you. And those are some of the largest AI applications in deployment on earth. We're huge in oil and gas, companies like Shell, where we're helping Shell reinvent itself, applying AI, upstream, downstream, midstream to fundamentally change the way that they do business. So that is the -- that's what we're up to. We're -- roughly today, we're running -- as of this quarter, I guess, we're running at about $0.25 billion run rate in revenue. I think most of the projections for this -- last quarter, we grew the top line 41% year-over-year. I think most of the projections for this year are that we will grow 35% to 36% year-over-year. So we're certainly in the top deck aisle. We're running basically a cash-neutral business, and the game that we're playing -- today, I believe that we are the world's leading provider of enterprise AI applications. And the game that we're playing is to see if we can establish and maintain a market leadership position in enterprise AI. This promises to be a huge addressable market. We have a rapidly growing business. And if we succeeded doing that, I believe we will have built one of the world's greatest software companies.

That's a fantastic overview. So I mean you're solving complex problems for some of the largest companies in the world. How do you define the addressable market, just given a lot of what you're describing is new? Are there things that you've seen from your existing customer base or ways that you think about the potential for certain verticals that lead to an estimate of the size of the market that you're going after here?

When we apply -- we look at the intersection of artificial intelligence and enterprise application software, it looks like basically an entire replacement market. I believe there is no enterprise application, be it CRM, be it ERP, be it supply chain demand management, customer management. There is no application that will not be predictive prospectively. I mean we simply will be applying AI models. We will be moving these applications into the cloud. Everybody is doing this in a manner of where they want to support a multi-cloud architecture, where it's a requirement that their applications run on not only Azure and AWS and the Google Cloud and the IBM Cloud and on-prem. So we have this kind of complex multi, hyperscaler cloud market that we need to run on. Those are the requirements of the market. If we look at the analysts' predictions for the size of the enterprise AI software market, this promises to be order of $300 billion software market in, say, 2025. So this will be the largest software market, most rapidly growing software market that I will have seen in my professional career. And so it's -- I think when it's all over, it's going to be a behemoth.

There's plenty to chew on if you're in the hundreds of billions, no question. When you think about the extensibility of what you've built and the extensibility of the platform, how much of it is the technology you've built and AI applications you're finding are similar across some of the key verticals that you're targeting versus once you land a target customer in oil and gas or financial services or one of those industries, there's repeatability there. And so you land with a flagship customer, and you're able to attract new customers that might have similar problems. Can you just talk about what you've learned and what allows you to expand both broadly and down certain verticals?

So when you can solve a problem, a very high-value problem, like AI-based predictive maintenance for offshore oil rigs for an oil and gas company, I mean, the value -- the cost of a failure is effectively infinite, right? And the -- and so when you can dramatically reduce the cost of that kind of catastrophic failure, when you can demonstrate significant social and economic benefit by allowing a company like Shell to deliver safer, cleaner, more reliable, renewable, lower-cost energy and where the economic benefit is measured in billions of euros a year, then others pay attention. So it -- when you can solve an anti-money laundering problem for a large financial institution, and a large financial institution, it's not uncommon that they're spending $1.5 billion a year just complying with AML regulations. ESG, huge; energy management, huge. So once you have a few [ showcase ] accounts in every one of these verticals, everybody else seems to want to follow. Now what's counterintuitive about what we've done, Michael, is we have -- basically 90% of -- we built something called a model-driven architecture, which is a very unique piece of technology that we patented and we hold all the -- it's a very unique piece of intellectual property. And what this enables us to do is, as counterintuitive as it may be, whether we're doing anti-money laundering in a bank, AI-based predictive maintenance for -- on an offshore oil rig, AI-enabled CRM or, say, stochastic optimization and supply chain for a manufacturer, 97% of the code that we're running in all those instances, it's the same code. All that varies are the data sources, the machine learning models and user experience. So that is really what's unique about what we've done, is the applicability of the code base to a wide range of problems, be it energy, precision medicine, manufacturing, aerospace, defense, financial services.

Got it. That's super helpful. So in terms of go-to-market then and how you reach these customers, how important is industry expertise in identifying these problems? Can you talk about the go-to-market resources you have and just what the sales cycle or sales motion might look like for landing a flagship customer?

Yes. So this is a -- clearly, this is a market share game. And so the -- just like it was a couple of decades ago in the relational database market, just like when we established the CRM market at Siebel Systems, I mean, this is a -- it's a market share game. So the game that we're playing is to see if we can -- the market it's in, it's nascent stages. I mean enterprise AI, we're in the first half of the first inning, okay? And we might just have the first guy up to bat. Actually, it might be that early in the game. And so we want to go out and gobble up as much market share as we possibly can. So we're organized geographically into North America, APAC, EMEA. Orthogonally to that, we're organized into vertical markets for financial services, oil and gas, energy, telecommunications, manufacturing. Associated with each vertical, we're tightly linked with a go-to-market partner. And for example, in oil and gas, we go to market, as you might know, with Baker Hughes. And so Baker Hughes is a large oil and gas service provider. And they have 12,000 people who -- selling with us around the world. In financial services, we go to market with FIS. In aerospace and defense, we go to market with Raytheon. So this is where we get the deep, deep industry expertise. Now again, orthogonally, both of those planes, if you can envision in this cube, okay, we go to market with the hyperscalers in a pretty big way. So we have a huge selling and service motion going on with Microsoft and -- our friends at Microsoft with their Azure Stack. And I think we've closed about -- in excess of a couple of hundred million dollars worth of business with them, and we have a very, very large pipeline that we're working. We go to market with Google Cloud. We go to market with AWS. We go to market with NVIDIA. So those all look like market partners. So that is -- so we're using a lot of market leverage to be able to expand into these markets very quickly.

Go-to-market is clear. That's a good response. And so when we think about the customer spend profile and just the way that you land within customers, is it more often that you find the targeted use case and then you're able to expand in other areas? Or are you landing more completely? And maybe you can just describe if there is a land-and-expand motion for investors, what that might look like.

I'd say it's definitely land and expand. So we have 40 turnkey enterprise applications that we offer today in the verticals that I described. And usually a company who wanted to solve -- they'll have a specific business use case that is a burning platform that they need to solve, be it supply network risk, which is getting increasingly important, be it AI-enabled CRM, be it cash management of the bank or anti-money laundering. And so once we go in and solve the first problem, then they seem to come up with 5, 10, 12, 17 other use cases and our footprint can get quite large.

Yes. So -- right. Because this is a financial conference, you touched on it earlier in the onset that you just printed Q2 results as well. So maybe you could characterize just some of the highlights and takeaways for investors from the financial side in Q2. And help us understand if there's a seasonal element to your business at all, just what that could look like.

I don't think there is any seasonality. Q2 was a pretty good quarter. I mean we -- top line growth was 41%. I think software revenue growth was 32%. We're ahead of our own -- of our -- of the estimates that we gave and the analyst expectations on virtually every metric. We're essentially cash neutral. We have about $1 billion cash in the bank. We're investing heavily in brand, brand equity, brand recognition and its sales, investing heavily and advancing the technology footprint. I think the -- if I were to look at the analyst estimates today, I believe they have us growing at 35% to 36% this year, up from 17% last year. So that 35%, 36%, that would mean we're roughly doubling every 2 years. And if we're able to continue to, right now, I believe that we should see modest increases in the growth rate to the next -- in the coming 3 years. And I expect this will be a large and successful in a steady state, cash positive, profitable, healthy enterprise.

That's great. When you think about the -- just all the opportunities you have in front of you, you described the size of the market opportunity, the potential to invest into R&D, some of the go-to-market efforts as well, how do you prioritize areas of investment? And when you're thinking about the balance between growth and margin, how do you sort of assess that framework and make sure you're appropriately focused in the longer term?

Well, clearly, our gross margins are going to be greater in the markets that are more mature. That being said, in order to grow in the long run, we're going to have to address every market, be it retail, consumer packaged goods, what have you. I'm particularly interested in applications related to defense and intelligence because when you look into that a little bit, cybersecurity, this is -- some of these exchanges that we're having with China and others in this area are existential in nature. And so the idea of having the opportunity to contribute to the security of the world is candidly highly motivating. The other area that I find of huge personal interest, and I am satisfied it will be the largest commercial application in artificial intelligence, is precision medicine. You could -- for example, it is within the state-of-the-art today to aggregate the health care records and the genome sequences of, say, name the country, maybe the United States, aggregate those data into a unified federated image, enabling us to use predictive analytics to identify what disease, who is likely to be diagnosed with what disease in the next 5 years so that we can intervene clinically and avoid the diagnosis. We will have genome-specific medical protocols that will be AI-driven, AI-assisted diagnoses and treatment plans. And then when you couple that type of technology with this kind of reach, then you can reach these historically unserved -- all right -- unserved...

Naturally.

Perfect timing, right? Yes. Unserved and underserved communities. So the social and economic consequences of that are going to be staggering.

It's all -- I mean, that's pretty -- those are pretty impressive and humungous problems that you're tackling. So when you're looking at the applications road map and the directions that you can go, how do you decide how many resources to allocate towards all of these really important problems?

Well, I would like to tell you that it's highly analytical of me. It's actually McKinsey guys around here [ unrolling ] spreadsheets and telling us which market opportunity to break next. But it's -- unfortunately, it's not quite that sophisticated. If the CEO of Boeing walks in here with his or her executive staff with a check and a mandate, we're in the aerospace business. Or the -- or if the Chairman or the Vice Chairman of the Joint Chiefs of Staff comes and spends the day here with the general officers of like every military agency and allows us to assist with work with the Missile Defense Agency or the Space Force, we're in the defense business. And so it's -- we're going into the markets as the markets are approaching us. And so it's coin operated, but I do think it's rational.

Yes. No, makes a lot of sense. It's hard to dispute that perspective. So one of the things that's come up in a series of conversations that, what you just touched on, is kind of interesting around as well is just the war for talent. Everyone's talking about just the age of employee empowerment and more emphasis on talent and some degree of labor shortage, especially in technical talent just given how much need there is there. And one of the things that a few companies have said is, "We think that it's not the biggest pay package necessarily. It's offering technical talent some of the most challenging problems in the world to help solve is what we're positioning ourselves at. Like we think that's ultimately what's going to keep people engaged. It's not just going to be a paycheck." And so in the spirit of that, given you've laid out some of the problems that are more complicated, maybe you can just talk about your ability to hire, retain talent, c3's positioning in the sort of tighter labor market for technical talent.

Well, first of all, I agree with everything that you've said. And I think that in -- particularly in our business, that will be the competitive advantage is. To the extent that we have a competitive advantage today, and I believe we do, it is the strength of our human capital. And the extent that we establish a leadership position in this market, it will be because of our human capital. Now I believe that we're extraordinarily fortunate in the quality of the human capital that we seem to attract. I believe we employ about 688 people today, rough numbers. In the last quarter, we had 18,000 job applicants. In the last year, we had 52,000 job applicants. And these are the best and the brightest out of MIT and Oxford, Harvard and -- okay? And can -- Verizon and JPMorgan Chase and -- you name it. And so we're -- in the last year, I think we had 52,000 job applicants. We interviewed 9,000 and hired 200. So we're extraordinarily, extraordinarily fortunate in the people that are raising their hands and saying, "Hey, I want to go to work for C3." And I don't -- they certainly don't want to go to C3 because we're all working from home because we're not working from home. We're here at work. I think they want to go work for C3 because we work on something really challenging, really exciting, something that's really high impact. The type of people we tend to attract like to work on extraordinarily difficult problems, and that's what we have: difficult, high-impact problems of enormous social and economic benefit that requires a lot of collaboration and teamwork to solve. So that's what we offer. And for that segment of the population, that is attracted to that kind of environment, we're a great place. For people who want to work from home, I mean, they should go to work for Facebook, okay? Sit in their pajamas or maybe, I guess, people now holding down 2 or 3 full-time jobs they understand. Well, they go work for Facebook, Google and [ Salesforce. ] And that's fine with them, but that's not the kind of people that we attract.

Yes. I see where you are right now, see the office setting behind you and beautiful offices in Redwood [ shores, ] by the way.

Thank you.

One of the other elements of just yourself, and I think what's always been -- what appears to be just a personal interest as well is just close ties with some leading universities and research efforts. So maybe we can step away from C3 and the business entirely and just talk a little bit about those efforts that you're helping on the research side with certain universities and if there are other sort of philanthropic areas of the business or just yourself that you'd highlight here.

I'm personally very active in the -- in higher education, particularly in North America and have been on the Boards of multiple schools at Berkeley and Stanford and Princeton and Illinois and others. And we enjoy at C3.ai a very close relationship with the academy. And we, for example, have sponsored a philanthropy, it now runs independently of us, called the C3.ai Digital Transformation Institute. And this is a collaborate effort that we put together with Microsoft, Berkeley, Stanford, Illinois, MIT, Princeton, Carnegie Mellon, KTH in Sweden, and I'm certain there's 1 or 2 that I forgot. But what we're doing is we're sponsoring a very, very advanced primary research in artificial intelligence. The first 3 projects that we have -- where we've held a call for papers, and I think we funded over 100 project -- research projects now. And by the way, all of the science goes into the public domain. So it's available to everybody. This had to do with developing new AI techniques for COVID and pandemic mitigation, understanding course of disease, drug discovery, efficacy of social mitigation. The second effort that we funded that was relatively large scale had to do with development of new AI and digital transformation techniques for energy and climate security. This is everything from building climate AI, for building climate models to carbon sequestration, integration of renewables, what have you. I think -- and the next effort that will be announced soon is going to be -- we'll be sponsoring quite a bit of advanced research on the development of new AI techniques for cybersecurity. The cybersecurity threat is getting very real. It's getting really troubling, and it is a natural application of AI to be able to shut that stat. So we enjoy a very, very active and fluid dialogue with the Academy. They are now, I think, holding bimonthly colloquy on advanced AI techniques. And this is not your usual kind of AI drivel. I mean these are like these are -- these are like Nobel laureate level people doing serious research. So it's really exciting. It's fun to be part of.

Yes. That's fantastic. You know you've done something right if you're allowed to work with both Cal and Stanford, by the way. Go Cardinal from my side, but that's pretty amazing.

Well, 2 great institutions.

Absolutely. A very spoiled rivalry.

You're a Stanford guy?

Yes, I did grad school there and -- special time.

Which school?

Management science and engineering. So all the stochastic stuff that you're talking about, that's been my wheelhouse, actually.

I'm on my way over to Stanford in an hour actually to go meet with [ Han Dee ] and the H.R. McMaster and others over at Hoover.

Wow, that's great. Enjoy the farm. Enjoy the farm. Hopefully, they have a few better places to eat over there, but that sounds like a great group. So we have time for just one more, and I want to turn it back to you and just in terms of closing thoughts, investors are focusing in on next year, durability of themes, hoping for some degree of normalization in the future. Your sort of parting thoughts around C3, your focus areas for 2022 and how you'll define success for yourself and the business over the coming years. What are the takeaways and conclusions from your side there?

It's very clear we're focused on top line growth. We're focused on market share. As we build -- I mean, what is the -- what's not to like about C3? I mean the scale of it is not sufficiently large that we've taken the lumpiness out of the quarter-to-quarter results. So we need to get the lumpiness out of the results. That will take -- that's going to take care of itself with scale and time. But we're focused on growth. We're focused on innovation, and we're focused on leading in our way a long string of highly satisfied customers. And if we succeeded in establishing a market leadership position in enterprise AI, this will be one of the important information technology companies. If we fall just short of that, it's going to be a hugely successful company. So that's what we're focused on, Michael.

That's a great closing thought. I just realized you don't have the suit and tie on for us. You have it on for your cohorts that you're meeting at Stanford who are much more important. But we appreciate you fitting us in, in the interim. It's a great conversation, and great to have you, Tom. I appreciate the time.

Thanks.