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

So look, we are just delighted to have Tom Siebel with us at The Citizens JMP Tech Conference. C3.ai is hosting their transform user events starting tomorrow. Where is that?

So it's our user group, International User Group Conference in Boca Raton. It started Wednesday morning at I think 8:00 East Coast time. And we will have our largest partners and customers speaking there about the -- I mean our product managers will get the product road map. There will be, I think, 500 or 600 customers and partners there. And then our customers, think Shell, think Dow, think Cargill, think United States Air Force, will be talking about what they -- their experiences of deploying the C3.ai solutions, the economical benefits what the benefit they realized and what they'd like to see from us going forward. So if you're interested in like firsthand, see them from the real people the extent to which they were receiving economic benefit from what we're doing and the extent to which they're being successful with these projects. I encourage you, you just go into our website, register and it will be livecast.

And my point on that is we're grateful that you made the time to come spend time with us, even though you've got all these people waiting for you in Florida. So thank you. The other thing that Tom pointed out to me on our call back, which I then bought is -- this is the tiniest book. This book is only 100 pages long and it's called what is ChatGPT doing? And why does it work? I'm on Page 2. okay? And I've already learned this is all I do, okay? So like -- and it is the clearest, best explanation -- so for example, does anyone -- can anyone to explain what temperature is in an LLM -- not you, Amit. Anyway, it doesn't work in C3, right? So by the time you get to Page 2, you will understand how it is that these models are adjusted to be more creative or less creative, and it is so simple. So anyways, this was Tom's recommendation, you can get on Amazon, it costs nothing and highly recommended. Okay, Tom, thank you for coming. We'll start out with my standard question, which this year is actually really great because you guys just reported your results, right, how's business?

Business is good. I think for some years now, we've been predicting that there would be an enterprise AI market. And while there was much speculation about the addressable size of that market over the last decade, whatever the speculation was. It turned out to be quite a bit larger than that. I looked up the market cap of public AI companies this morning on my way in on Google Bard of all things, in the combined market cap of public AI companies is somewhere between $7 trillion and $9 trillion. So it appears that there might be a market here. And we're -- we spent the last 15 years getting ready for this. We spent over a decade and a couple of billion dollars building a platform that enables a software platform that allows us to design, develop provision and operate large-scale enterprise AI applications. On top of that, we've built more than 40 turnkey enterprise AI applications. So as we enter March of 2024, the market seems to be coming our way. And a lot of people want to deploy enterprise AI applications. And there is, I think, one company in the world that offers a family of enterprise applications that address the value chains, financial services, defense, intelligence, oil and gas, manufacturing, what have you, and that would be C3.ai. So business looks promising. We've made it through. Our pricing transition from subscription-based pricing to consumption-based pricing. We experienced the downturn in growth rates that we predicted, okay? And now we are experiencing the upturn in growth rates that we also predicted. And for those of you who still know how to do progress your models, you'll find that the correlation between our growth rate and the stock price approach is one. And so as we return to rapid growth, the stock price seems to be well correlated with that.

Yes, when you accelerate stocks tend to go up. So talk to us a little bit what is the demand like? What are the conversations that you're having like in the federal government? What's it like in enterprises here in the U.S. and then what's it like internationally?

There's no secretary of a service, no government leader, no CEO in the world who will not take a meeting about enterprise AI today, not one, okay? And so the doors are wide open, and they're trying to figure out what it is, how it works, the way that we have the product price today, where it's -- we will take a value -- a supply chain and demand chain or a production line, and we'll optimize that using -- we'll bring that into production with an enterprise AI application and full production years in 6 months for $0.5 million -- $0.5 million. I mean this person doesn't need budget, okay? It doesn't issue an RFP. This version just signs it for $0.5 million. It's live in 6 months. And then if they like it, they keep it. And so we're finding the interest in what we do is very high. The number of customers with which we're doing business has increased dramatically. I think 85% increase in customer engagement year-over-year, if I'm not mistaken. Okay. And business is good. I think we have 40 -- 45 customer engagements today, up 85% year-over-year. The market for what we're doing is on fire.

Yes. And most of the companies that we talked to you don't actually talk about customer engagement and measure it the way you do. So explain to us how do you define it?

So it's difficult for us to identify Department of Defense, for example, as a customer, okay? For example, we have -- I have 3 different projects in the Missile Defense Agency, and those are unique budgets, unique projects, okay? And those are 3 customer engagements. I have multiple engagements in the United States Air Force, and those are unique customer -- separate budget, separate problem, okay, a separate division of the Air Force. And so those would be -- that would be a customer engagement. In an organization like Koch Industries were installed at Flint Hills, resource is clearly quite different than Georgia Pacific. It's just quite different than Molex, and those would be different customer engagements at Koch.

All right. Great. So for people who don't know, so this business in 2022 grew basically 40% then 2023 decelerated to 6% and now is starting to reaccelerate again. What drove that? And where can it go?

Well, I think we're so in Q2 of '22, we're growing at 42% exactly, okay? We announced the transition from subscription-based pricing, which meant our transactions were $10 million, $20 million, $30 million, $40 million, $50 million upfront to get started. This is how we operated as a public -- as a private company. And so our customers finance the business, and we would do business at $39 million or EUR 39 million all day long, which is good work if you can get it, you don't have to ring a doorbell on the Sand Hill Road. Okay, as we transition to a public company, we transition to consumption-based pricing, which is just pay as you go. Ipso facto, we saw a decline in our growth rates from 42% year-over-year growth rates to 0. And now as consumption pricing has kicked in, it's increased from 0% to 11% to 17% to 18%. We've seen dramatic growth in underlying infrastructure in AI in the past year or so. I think NVIDIA revenue was grew up by 260% or so year-over-year, GPU revenue up 400% year-over-year. What we're seeing happening with the cloud providers is quite dramatic. I think the enterprise value of public AI companies today is $7 trillion. And so this turned out -- this AI market turned out to be the real deal. We play in an important segment of it and the game that we're playing is to see if we're going to establish and maintain a market leadership position in enterprise AI, okay? This is not social media. This is not deep fakes. This is running core business processes, supply chain, demand chain, demand forecasting fraud, a manufacturing optimization customer churn customer satisfaction, et cetera. And that is a large and rapidly growing market. Now at Oracle, when we started that, we were nothing, and we said we're going to establish and maintain a market leadership position in relational database. I think we did a pretty good job at Siebel, a decade later, Siebel Systems, we set out to establish and maintain a market leadership position in CRM. We created that market, and we sold -- when we sold the company to Oracle, we had 85% market share. I think our revenue at the time was order of magnitude a couple of billion and sales force was order of magnitude a couple of hundred million. So I mean, the company was doing pretty well. And after that, Mark did a pretty good job of building his business, so good for him. The -- we might not succeed we might fail. We may end up #2 or #3 okay? And the #2 or #3 company in enterprise AI, okay, is not worth $3 billion.

Currently, you're guiding around a little over $300 million in revenue, right, for '24?

For '24, yes, I think that's right. You know better than me.

Order of magnitude, you're talking about -- it's still a really small business in the grand scheme of things, yes. How big can it get?

It sure is. It's a miniscule. I think this is -- can be one of the largest and most successful companies in the information technology business. I do. I think this is the real deal. We have great products. We have great customers, listening to the -- dialing of the users group meaning Wednesday or Thursday or have whatever your people do it and see what the customers say. I mean customers will stand up and say there's a customer who will stand up. Shell will stand up and say they have stood up and said they're in $2 billion in economic benefit a year from using our products. People, how many people will sit up at SAP a conference, an Oracle Conference, a Siebel conference or a Salesforce conference, an IBM conference, okay, and said they're getting $2 billion in economic benefit a year from their products? That would be 0. And so this is -- I encourage you to dial in and see what they have to say.

Yes. And you took the business from losing money to basically to getting to breakeven. And then this last quarter, I think, operating loss or the operating cash flow was negative $39 million. So what was the...

You're talking about breakeven in cash flow because I don't think we've made a breakeven in profitability yet, Pat. I think that happens next year. And so we will run -- at Siebel, I ran a consistently cash positive profitable business for a long time. I am a cash positive, okay, profitable business guy. This is what I do. Okay. I do really understand that I get it, okay? And we were running a cash positive profitable business back in the dot-com era when that was completely out of vogue, okay? So we're -- so our goal is to run a consistently cash positive non-GAAP profitable business. okay? That's the business plan. We will do that, and I have some experience and track record doing -- it's not that hard, guys. I mean, if you have a structurally profitable business, which we do, what do I mean by structurally profitable. If your cost of sales and your cost of selling is lower than your revenue, you have a structurally profitable business, right? We can turn this into profitability tomorrow. All I got to do allow some people and cut some marketing and R&D costs. And it's -- again, it's non-GAAP profitable and cash positive model. Is that in the best interest of the shareholders of C3? No way I know how, okay? But so this business of running a profitable business. I mean, it's not that hard. I mean, I didn't get to go to the Harvard Business School, but I imagine if they had a class on that, it would be something like figure out how much revenue you all come in and spend less than that, got it. Okay. I really do understand that, and this is what we will do.

But you decided to reinvest, is the point I was making.

We are reinvesting. I am investing in generative AI.

Explain to us when you made that decision and why you did it?

I made that decision, I think about 4 months ago. And this -- when we got into this generative AI market, which emerged really as we got into 2023. And every time we look at it, it's bigger than we thought. It's -- I have no idea how big this is. It's just breathtaking. And the -- we've come up with a unique, highly differentiated solution. We continue to be really impressed by the creativity through which people are applying this technology but we are investing in generative AI. We're establishing brand. We're establishing brand leadership. We are building products. We're establishing market share. And if you -- I gave the leading indicators in terms of what we're seeing in terms of web traffic, lead traffic, what have you, and they're extraordinarily positive. But we have, I think, $750 million in the bank. So we're not -- I can assure you, for those who have -- people who come to my desk and want to spend money, the default answer is no. We start with no, okay? And then we move from there. So we're very, I think, discriminating how we spend money, but we're a well-capitalized business, and we're investing it prudently into this to establish a market share or established market share in generativen AI.

Okay. One more for me and then we'll open it up to the audience, and I totally didn't include this on the list of questions, sorry. The -- what do you make of Elon Musk lawsuit? Did you get a chance to look at that where he's suing open AI and he's basically saying, hey look...

It sounds to me like he's dead right. I mean, well, those guys had to meet all in. They had the meeting at Madera, okay. And what was it 2015 with Brockman and with Greg Brockman and what's his name Altman. And these guys decided that it was -- Google had all the human capital and all the Google stack was proprietary. And the deal was that they were going to build the deal -- back then, they were going after AGI, which I would argue, LLMs are not. But okay, that was the goal to go after Artificial General Intelligence, you know what all but you know what that is. Okay. And Elon said, listen, I'll write a check up to whatever. You guys go raise the money and whatever you don't get, I'll throw it up to $1 billion. I mean he did it, okay? And the deal was everything we're doing is going to be open source, okay, and everything we're going to do is going to be free. That was the deal.

And for the benefit of humanity, yeah.

That was the bargain. And so I think he's got a legitimate gripe, whether anything -- whether there is a rule of justice in the United States, I'm not quite unsure and how that might work out in the court system, I don't know. But we may if you play back the tapes on this, that was the bargain. And that's definitely not what they're doing. It is not for the benefit of mankind. It is not open source, and it's not available to everybody. That was the deal. So he's got a legitimate gripe, but he did pay for it.

Questions from the audience?

If you please say -- identify yourself, please.

[indiscernible] Okay. Fast forward 3 to 5 years and imagine you are saying to your C-suite staff -- oh c***, I didn't see that coming? What do I need you talking about?

Well, I'd tell you I think that the constraint will be in 3 to 5 years. I think it will be power, okay. Okay. We believe today -- I mean, right now, we have all these guys -- my friend, Jensen, okay, and the people and Andy Jazz and others kind of laying the infrastructure for what we do. okay, with all these -- the cloud infrastructure and all these GPUs, okay? I think the constraint is not going to be the availability of GPUs people -- as they believe today, I think it will be availability of power. Okay, see PG&E for details. So if we want to build a data center, in Redwood City, PG&E will not provide us power, okay? And as you go south -- south silicon value, you also cannot get power, okay? So I think power becomes the -- at first, it was change management, then it was GPUs that appeared to be the constraint. Well, first, it was software. We got beyond all that. I think the constraint will be availability of power. And that's critical.

So here's a great data point I heard on that from another CEO who's here, which is that to get 1 lousy megawatt of power from the leading provider in Japan, he has to wait until Q2 2025.

That would be TEPCO.

1 megawatt, yes.

Energy efficient industrial nation on the planet. I mean, those guys are really good. Other questions?

My name is Robin, I'm with the Investment Strategy Analysis. I have a question. Do you think we are in an AI bubble of sorts or is this just the beginning of a giant one?

Either -- I don't know if we're in a bubble. We will be in an AI bubble at some point -- look at every -- if we look at every major technology change from transcontinental railroads, urban electrification, the Internet. I mean they all result in the market all kind of the overvalued initially, these technology changes, which were very significant positive technology changes. And let's go look at the dot-com explosion. Many of you were there, and I can assure you I was there. Okay. And the -- and again, so this thing blew up about 2000, okay? How big was the Internet in 2000? How big is the Internet now like maybe 3 orders of magnitude bigger today. So however big we thought it might be then? I mean, we couldn't have imagined what it is today. So will the market correct itself at some -- Will it overvalue this? Yes. Will it correct at some point? Yes, is what markets do. And here's why you guys make the big boxes to time that correction. Good news is I don't have to worry about that. And I have to focus on making customers satisfied and growing the business, growing the business, growing the business. And the market correction will correct, and then it'll take care of itself. Good thankful, I don't have your jobs. You guys have hard jobs. Other questions? Looks like we're off the hook, Pat. So what are we focused on at C3. We're focused on technology leadership, and we're focused on thought leadership. Okay? And as we move into '25, '24, we're going to be questioning '24, '25, '26 we're focused on growth, growth, growth. We focus on market share, okay? In the way these markets work, somebody is going to establish a market leadership position, if not us, who? Okay. And the -- so we're focused on accelerating the market and establishing market-leading position. Question?

Austin, CJMP. Question is the 15 years you've been developing this technology. Can you break down for us what in those 15 years, giving you kind of the edge? Are there any other players that are coming into this space to talk about the expanding market, but you still have to be vigilant in making sure that you are tracking with that experience?

Great question, Austin. I would say the investment we made beginning in the first decade was the most important one. So we understand when we started, AWS had just gotten started. There was no Azure. There was no Google Cloud, right? Nobody talked about enterprise AI. Nobody talked about enterprise AI -- and people kind of remember Marvin Minsky talking about it in Princeton in 1951, but that's all they knew okay? Now the -- I think the investment we made that differentiates us was the 10- to 15-year investment we made in the C3.ai platform. The software stack that just massively differentiates us from everybody else in the market. So what Watson -- what our -- actually, what IBM argued they built with Watson, which was a complete fraud, we built it, okay? With GE attempted to build with Predix, we build it Okay. And so whether we're deploying what's really a counterintuitive about this, whether I'm deploying predictive maintenance at the scale of the United States Air Force, customer churn in Bank of America, whether I'm doing stochastic optimization of just demand chain and supply chain at Cargill 90% percent of the code predict demand for offshore hovering, hydrocarbon loss accounting in Shell. -- it's 90% of the code is identical across all those deployments. So the C3.ai platform is a core differentiator that distinguishes us from any other attempts to enter this market.

All right. I'll ask the last 1 in the last 30 seconds. Biggest near-term opportunity for C3 in the federal government?

Biggest near-term opportunity in the federal government. We have an application called CBM+ from the Rapid Sustainment Office, where we are the standard for predictive maintenance in the Department of the Air Force. So one of the assets that we -- that where we work is in aircraft, and the United States Air Force has 5,000 aircraft. And in a given day, about half of those will deploy and half are down for unscheduled maintenance. We can use AI to predict system failure and subsystem failure before it happens. So auxiliary power unit, flap actuator, Igniter and the Afterburner of B-1 Bomber or whatever it might be. If you can predict it you can identify the failure of 50 to 100 flat hours before it happens, you can dispatch the personnel, the material, the converge with the aircraft, fix it that night and meaning it flies away. And net-net, we get 25% increase in aircraft availability. Biggest short-term opportunity is to move this from the Air Force to the Navy. The Navy has more planes than the Air Force to the United States, armory to the Marines, okay? So that's the biggest defense opportunity, this opportunity in the short run is actually state and local government. I never anticipated this, and this is turning into a really...

Tom, we went over. So the short version on the state and local government?

There are 4 applications that we're dealing with. One is we brought AI to the assessment problem, property assessment. Why is property assessment important at the county level because this is the only source of their revenue. So we decreased the time and do an assessment of a commercial or a real estate structure by like a factor of 10. And the other area that's really turning huge is -- this is not an AI application, by the way, but the pendulum has swung and even right here in River City in San Francisco, law enforcement is coming back into vogue. So we have a C3.ai law enforcement app that we're deploying it, law enforcement agencies all across the United States that is really gaining a lot of traction.

Thank you. Well, thank you so much. Great having you here. We really appreciate it.

Thanks, Pat.

Good to see you.