Okay. So, you guys are lucky enough to have the team of Tony Rodriquez as the CFO of Cerence, and I'm Brian Krzanich, the CEO of Cerence. We're here at the Needham Growth Conference. I'm not gonna go through this, but if there were any forward-looking statements, just remember the caveat of their projections, their forecasts, and you can't hold us accountable. Just like you guys can never be held accountable for the information you give to the investors. I'm joking. I thought today we're gonna give just an overview, but we're gonna. I didn't wanna do the classic, like, "Oh, here's who we are," and there's a little bit of that in here, but what I really wanted to do is actually answer a question that I got a lot at CES from people. It was, "Why Voice? Why, why are you guys in business?
“Why, why is Voice? And why do you think it's becoming very important?” And I think it's going through a transition voice right now, and so I thought I would answer that question first, and then we'll spend a little time about what the company is and all that, so. LLMs plus Voice is what's generating a great user experience. And that's the difference between Voice five years ago and Voice today. Is that these large language models have completely transitioned what Voice is, what it can do, and what it means to you as a consumer, and what it means to us as a company too, quite a bit. So we're gonna explain why. I see that inquisitive look in your face. It no longer requires keywords. That's the fundamental difference. You can speak naturally.
So if you go back to voice, I have a 2016 Ford F-250. To get anything done through voice on that vehicle, I have to say exactly the right word or, or right set of words. If I say, "Turn on Bluetooth," it doesn't do things. If I say, "Bluetooth on," it connects to my phone, right? That's the old voice. What it is today is that it allows you to just speak naturally. So it eliminates the need for you to take your eyes off the road. So, why have voice? Why is it becoming so important? The, the first is that it's, it's actually generating a more safe environment. You're not reaching down to your phone. Our vision is that you get in your car and you put your phone down, and you don't have to ever pick it up.
That if you don't wanna touch a knob in your car, you don't ever have to touch a knob in your car. That's our vision. And we're there, by the way. We, we can do this already. And the cars that are gonna be coming out in 2026, you can do multiple instructions in one long sentence. So no longer do I have to say, "Bluetooth on, pause, play." That's how my Ford F-250 works. I can say, "Hey, Cerence, can you crack the window? My feet are cold, and can you find me a Starbucks along our route?" Just speak like that. And I can do that today. Let me tell you the difference in our—my life. My wife's Mercedes, little SUV, she can do this. You can already do this. Crack the window. My feet are cold.
I can say, "My feet are cold," and it says, "Okay," and what the system will respond right now is, "I'm cracking the window. I'm turning on the floor heat, and I found three Starbucks along your route. Which one would you like to do?" That's exactly how the system will respond to this request. That's the difference, and that's what LLMs allow, is that natural language. It now interprets these and breaks these into individual instructions and tokens, as we like to call them in the AI world, and the last thing is that Voice gives you simplicity, speed, and safety, so you can sit there in your car and say, again, not taking your eyes off the road, "It's snowing out. Can you increase the distance from the car in front? And please turn on the seat heat," right?
It will adjust now the Adaptive Cruise Control, give you more distance between, because usually within Adaptive Cruise Control, there's a set of settings that if you go about five buttons into your system, you can adjust the distance between you and the car in front. The system will automatically do that. But rather than going five buttons into your screens, and hopefully remembering where that is even on your screens, I bet you most of you guys don't know where that, how to get to that screen, in your car, you can just say it, and the system will do it. If it's snowing out, all of that can be done very simply. Now, why do you care and why do I care? So I went and grabbed some data. Right now, about 40% of cars are connected.
it. I have to say about because, you know, the Chinese don't necessarily give out great data, and, you know, there's cars all over the place, and there's not good records and things like that. The estimate is. I've actually seen anywhere from. Now I'm seeing more. The number's getting higher. 80%-95% will be connected by the end of this decade. So the connection allows now also things like real-time information. We call it real-time knowledge. Who won the football game last night? Is the Steelers coach still employed? I watched the game last night, and I don't. I think his job's on the line now. Connection really gives you a lot more capability with these LLMs, right? There's about $400-almost $500 of annual revenue that's attainable, they believe, through the car by connections.
That's gonna be things like you're doing restaurant ordering, you're doing parking requests, you know, "Hey, can you tell me what's parking? And can I pay in advance?" Things like that. 80% of car buyers want embedded AI. So when we think about AI, we work on it at two levels. We call embedded that sits on the vehicle itself. There's enough SoC capability. It's usually a mid-level Qualcomm MediaTek ARM SoC and memory. We want four to seven gigabytes of memory on the vehicle, which is not a lot if you guys know memory very well, to be able to have a good embedded experience. So that's an S—we call it an SLM, a small language model. And we're putting more and more capability onto that small language model so that you can do more and more without being connected.
You may be in an environment like in some of these city streets here, in a tunnel, whatever, and you don't have a connection, but you still wanna be able to change that adaptive cruise control. You may still wanna adjust your mapping locations, find points of interest, and the connection's just not there for whatever reason, even though your car's connected. There's a lot of parts of the world where there is no connection. There's just the cars aren't connected, and there's not good 5G like we're used to. 80% of drivers want personalized infotainment. They want that, "Hey, can you play my favorite songs? Can you play this? Can you do that?" This really allows us to do that. You know, offline performance is key. This embedded is critical. We really lead in that. That's how we compete against guys like Google.
Google really wants to push everybody to the cloud, and we wanna say, "We don't care. Cloud, embedded, we wanna be good at both." Whatever you want as an OEM and an end user will do. Then there's declining brand loyalty, and so OEMs are looking for this for how do they personalize the experience in your vehicle. How do they make the Audi experience or the Mercedes experience genuinely unique? That's another place where we compete against Google. Google wants it to be everything generic and only use Gemini. We'll help with things like, if you go to look at a Renault vehicle, there's a little avatar called Renault, and we sync our voice with the little avatar, and he's your personal assistant on the vehicle.
We are helping OEMs really make this personalization come back, and give you that experience. So the consumer's demanding this. We believe that it's a much better experience, and that's why Voice, okay? This is just a quick thing about Cerence. We really are in automotive, but we're really driving more and more to automotive and beyond. You know, we're in most every single OEM. We have a lot of data that we train against that's unique to automotive. This is a key feature here because a lot of people may have a lot of data, but it's not automotive expertise. It's not automotive data. And so you know, we know what these words mean. We're in about 52% of the vehicles.
Another one that's really valuable to us is the number of languages that we support, you know, and we have to be able to support all the variations of language, right? A German person speaking Spanish doesn't sound the same as a Spanish person speaking Spanish, and we have to be able to interpret those words, right, and so we're really pushing across this. We're in automotive. We're in two-wheelers. We're getting more and more into autonomous vehicles. We're not building ADAS. We're not gonna do anything there. It's really just the speech portion of this. We're in trucking already. We're in wearables on Garmin, and we're starting to push into some of these other areas, okay? We're in LG TVs as an example. We just won a lawsuit against Samsung, which is mostly about TVs. There's our team. There's Tony. There, right?
But we have really strong, with Nils and Christian. They've been in the automotive world for almost their whole careers at companies, sometimes at the OEMs like Mercedes and some others. And they may have been at Nuance even before, like Christian. They went to Amazon. And then I've worked in and around automotive for probably 20 years. This is just a brand chart to give you a feeling for we're in just about every brand at some level. And it just depends on how much of our software they want to absorb and take. And even, you know, if you take a look at these models and model years, we can have all different kinds of varying levels of capability. Okay. So our vision, like I said, is we want a human-like interaction. We wanna be proactive and personalized. We're working with OEMs.
You get in the car, there could be facial recognition in with using the cameras inside. We see that it's you. It welcomes you, says, "Okay, it's 8:00 A.M. Are we going to work today?" Usually, you like to stop at Starbucks. Do you want us to stop there? Whatever that personalization experience is. We have a lot of examples of that that we're working on. We're really, you know, we call it, agnostic, but we, we wanna provide whatever the OEM wants to provide as that experience. And they're getting more and more creative in what that experience is. So, like I said, you know, Renault has the avatar. Some of the Chinese are really producing, some really unique capabilities in their vehicles. And, and then we believe that we are creating this Gen AI that's really tailored for automotive. Kind of already gone through most of this.
I wanna keep us going to get a lot of questions. We really think about the world in kind of two ways a bit. This conversational and generative AI is a lot of what we've talked about this morning, which is really focused on, you know, the experience and the words you can use and what you can do in the vehicle. But we also have a lot of knowledge about audio capabilities. And if you came by our booth at CES, we had some examples that we showed you. But for example, we're working with OEMs right now. We have the capability to really filter out various voices.
So, you know, one of the most common experiences, especially if you have a large family and kids and things like that, is you're driving along and, you know, the radio's going and the kids are, you know, either singing along or whatever they're talking and having a big conversation, and you get an incoming call, and you have to either silence everybody and turn off the radio or whatever, maybe try and put your earbuds on really quick. But with our system now, we can isolate everybody. We can isolate every single person in that car by their seat and take out all the noise. So you can be on your, you could be driving and you could be on your phone call, and you only hear the phone and they only hear you while the radio's playing to everybody else and they're talking at a normal voice, okay?
That is functioning, working, capable. So we have a lot of audio expertise that allows that kind of capability inside vehicles. And then, we just have a whole portfolio. I was just trying to show you there's a large collection of products that we do now. I wanna get to, we have a lot of, we believe there's a lot of opportunities beyond the car. You see our cross-licensing agreement that we did with Samsung for $50 million roughly. We have similar actions going on with other companies, right now. And there's quite a bit of work going on right now in things like self-service kiosks, other televisions. Like I told you, we were already at LG TVs, and we're starting to, we have a lot of car knowledge.
We've already ingested most of the owner's manuals of vehicles, because we wanna be able to do things like turn on the Adaptive Cruise Control or adjust it. So we can help with car dealers to take a lot of their call volume away and things like that. So, we've been working on products like that. At CES, we demonstrated many examples of this. And then I'm gonna hand it over to Tony. He's gonna talk to you through really quickly probably what you're really here for, the financial results.
All right. Excuse me.
Oh, sorry. There you go. All right. So we just have a couple slides on our financials. One, we're September 30 year-end, so our Q4 was for September 30 of last year. And so obviously, December 31 was our first quarter. We haven't released our results for our first quarter.
So we'll talk a little bit about Q4 fiscal 2025 and then what that looks like for next year. When Brian and I, Brian and I are now both in our second year with the business. I'm about four months, I think, ahead of Brian and in my tenure. But what we said was we want to, you know, we're, you know, if you follow the company, we're deep into the automotive industry. And so we said at this stage of this life cycle, we should be growing, but increasingly profitable, increasingly cash flow flowing. So, as we look at our Q4 of 2025 in the box, we did roughly $60 million of revenue at a margin of 73% roughly. But important numbers, the Adjusted EBITDA at $8 million and cash flow from operations at $12 million.
So as we put that in perspective of the year, we did about $252 million of revenue. Again, around that same 73% gross margins, we did $48 million of EBITDA and $46 million of free cash flow. The importance of those numbers is that the company wasn't previously cash flowing positive on the general business and profitable. So we did a big restructuring at the end of this year, Q4 of 2024. And you can obviously see the growth, and you know, the fact that adjusted EBITDA, even with a lot of the restructuring costs taken out, we were still negative in EBITDA last year, but we've been able to you know grow the business in a profitable manner in 2025. As we end in over achieving in guidance.
Those numbers on guidance were for the quarter, but when we entered the year, we were able to tell the investors, "Look, we're going to be. Our core technology will be growing. We're gonna have, you know, positive Adjusted EBITDA and positive cash flow," and we exceeded on all those metrics in 2025. As we think about, and let's just focus for a second on the full year. From the $252 million in revenue, our guidance right now for the full fiscal 2026 is $300 million to $320 million, higher gross margin. And if you look at the Adjusted EBITDA, you know, again, I'll use mid-ranges for both cash and EBITDA, Free Cash Flow. Roughly $60 million of Adjusted EBITDA and $60 million of Free Cash Flow. One small subtlety to what Brian said with Samsung, so it and why it's important.
We didn't actually go to court with Samsung. We were in a process of. They were in violation of our IP, and we worked with them all the way up to the doorsteps of the court. Then ultimately what we did, instead of not settling, but we sold them a patent license for their handhelds in the US. So we were able to sell a patent license for $49 million. That's in our Q1 period. You can see that Q1 we're looking at 110-120 in revenue. That includes a patent license of $49 million for that quarter. We've not projected any additional IP monetization revenue the rest of the year in that 300 number. What it really means is that our core technology is growing. Core technology revenue is growing about high single digits%.
And if you then fold in the IP monetization revenue, we're growing the business 23% at the mid-range, with, again, increasing EBITDA and increasing cash flow. So all positive metrics, which are really showing the benefits of what we've been trying to do, with the business. And again, last thing on the Samsung is, as Brian mentioned, we're, you know, we're primarily in automotive. Our goal was to really fix the core business, get it growing, profitable, cash flowing, and then look at what we can do outside of automotive. The way you should look at the Samsung IP license at $49 million is that was for non-automotive revenue. And we have, you know, a number of cases that are just now in backlog that will bring our business development folks to those companies to try and do the same thing.
How do we monetize our technology outside of automotive? The next two are in ITC court right now, Sony and TCL for televisions. So we expect additional IP monetization, but for the fiscal 2026 in that $300-$320 million, we've only included what we've already sold, which is the Samsung patent license. Okay. So that's it for our slides. Or if you're old, you're foiled. And we can have some time. We have about five minutes, five and a half minutes for questions if you have any.
Yeah. I noticed I saw Nuance is on the list and.
Yes. Spun out of the. Who's Nuance?
So how does what is going on there with, is there an IP infringement or?
In that case, it's kind of an IP infringement and contract.
We had a contract with them and Microsoft, so Microsoft bought Nuance. Yeah. Right? The only reason we, those two are really the same. Yeah. In here on this one. Should we swap between the two of those? Yeah. And we should probably edit our slide for that. There was a contract that they got access to some of our IP. The contract expired. They continued to use our IP and weren't willing to negotiate a contract, just wanted to keep using it. Interesting. Yeah, so that one's really more business contract related. You might wanna say that we're still working with Microsoft. You wanna talk about the, the. Yeah. And so that's what that one is. It's an IP disagreement on contract, right? And so we think they still owe us for that contract.
It you know, this kind of stuff, whether it's Samsung or Apple or Microsoft, happens all the time, you know, disagreements and, "Hey, we think you're infringing and, and all." But it doesn't mean you can't do business together. So we're doing a great work at CES. We had in the vehicle our full Microsoft Office 365 suite running. So you can now, the vehicle's now a trusted device. Yeah. And so you can get your corporate email and your full Outlook suite in the vehicle. So even though this thing is going on, it's just a contract. It's you know, not a large number of money, but we continue to work really well with Microsoft as an example. Same with Apple, same with Samsung. This stuff happens, right?
We think the company in the past wasn't very diligent about really defending its IP and being more forceful about it. And so we're just going back, you know, the Samsung one is an example where, you know, these are fundamental patents, things like wake-up word and the real fundamentals of speech to text. And you know, they should have been a little more aggressive about, "Hey, you can't do that with, we have these patents." And so a lot of this is we're going back and saying, "Okay, well, this is what you owe us for what you've used, and now let's talk about what it costs for the future." So Samsung has a license now to use our technology, as an example. That's what they get with that. So it's kind of a cleanup of the past.
That's what we're going through and doing on a lot of these. We always start with, "By the way, we, we don't wanna do this, and we'd rather just make a business agreement." That's where we start with all of these. There's dozens of letters going out now saying, "Hey, we, we'd really like to do a business agreement." Okay. Yes, sir. Can you remind us, do you get paid on auto production or sales? So it depends. Yeah, I could say yes. How about that? There's the answer. We think of the, remember I said we think of our stack in two levels, embedded, which is what sits on the vehicle, and the connected, which is, you know, a 5G or 3G connection. The embedded portion, we get paid when the car ships from the factory. Okay.
Tony gets a check in the mail from, you know, whoever, Volkswagen, Honda, whatever, every car that's produced. We do a lot of, a lot of the things we do in the during the quarter is correlate between, you know, vehicle production and what did they say we they built and what do we get paid, right? 'Cause it's not always real clean 'cause they have factories all over the world, and they're usually scrambling at the end of the quarter to figure out what they built. The connected portion, so the part that's the the 5G connection, we get paid when the vehicle leaves the dealer's lot. So when you drive it off as a consumer, that's when my pay, and that's a contract over time. The embedded is a you're, we've sold you the vehicle, you get to use that embedded.
The connected is a contract for a period of time. In my tenure, I've seen anywhere from one to 12 years, the different contracts that have come through. We say the average is three. My guess is in a year, we'll come back and tell you the now average is probably moving a little bit higher, four or five, 'cause we're seeing more of 'em. They're like, "Hey, this is really important, and we just want it," so Tony gets paid when that one leaves the dealership, but he has to take that revenue and spread it out over the length of that contract, so you'll see our billings have one number in our revenue that we get to actually declare is a different number, and that's important to note too.
That's a good question because if you see our revenue streams on our P&L, we will have connected, licensing, connected, and professional services. And that connected piece is actually recurring revenue. So, we get a billing when the car gets sold. We amortize that over a subscription period. And if the consumer likes that technology, obviously at some point they'll re-up with the OEM for that, and the OEM owes us then as a renewal for that as well. So that middle line item is a recurring revenue stream. And that is the fastest growing segment of our revenue right now. It grew double digits last quarter or last year. And what gives you the confidence that it's stretching from the average of three to? Just what I'm seeing. Yeah. I mean, it's just empirical.
Like, one of 'em we're bidding for right now is 12 years. That they want an agreement, a license for 12 years. In that case, the OEMs like, "Well, that's probably the life of the vehicle. So let's negotiate that you'll have this feature as connected feature for 12 years, and we'll pay you all upfront for those 12 years." Now, if you average it out on a per year basis, it's a lower per year, so they get a little break on that. But yeah, exactly. So but yeah, but it's for a cash flow standpoint, of course. One of the things you learn if you work in this industry around automotive is they often don't know what the heck's going on with their vehicle suite.
So their ability to track which vehicles are coming up on their licenses and, "Am I going out and contacting consumers and getting a renewal?" They're just now starting to figure that out, and they're realizing that's gonna be a lot of work. So a lot of 'em are coming back now and just saying, "You know what? I don't wanna deal with you. I gotta go figure out how to deal with the consumer and get paid. And maybe I wanna add a bunch of other things along with that, service agreements or whatever, right?" and so I'd rather just pay you 'cause for the length of this thing, and so I've got that out of my way, and I'm just gonna deal with the consumer.
And I'm gonna try the, what I'm seeing a lot of 'em do is try and add additional things onto those renewals. So we'll see. The renewals are just kind of starting, and I'd tell you the early numbers aren't very good because they didn't have LLM, so like, like I'm, I wouldn't renew my Ford F-250 connectivity. This piece of, you know, it's a piece of crap. I'm just gonna be honest, right, and so I, I wouldn't renew it, but my wife's Mercedes, that's two years old, if it goes up, I'm gonna renew it because it's pretty cool, so I think, you know, but that one isn't up for renewal probably for a year or two, so we're, we're still in kind of that transition between LLMs and connectivity that, so we don't even know what the renewal rates are gonna be yet.
That's why we don't project any renewal in our forecasts. Yeah. We just take the initial contract and because we just don't know. So as if we start to see good numbers and we could build a model, we'll start forecasting it and we'll talk to you about it. Yeah. What type of visibility do you give your guidance? What type of visibility do you have in a year? 'Cause I feel like the company historically had so many fits and starts. Yeah. So there's a couple reasons we have pretty good visibility. I'm always hesitant because, you know, I could be fired tomorrow 'cause I screwed something up visibility-wise.
But you know, what's different about the company is we don't do. The company used to do up to like $70 million a year of Pull Ahead fixed contracts, fixed contracts, right? Where they'd ask for an OEM to pay in advance for, you know, vehicles that are gonna ship for the next two years. And so they always had this snowball that they were trying to replace. And so they were always projecting that they would replace the snowball. We don't. We took that from the roughly $70 million run rate down to about $20 million. And the only reason we even, I'd have it zero if I could, but there's some OEMs who really want that. It's that same thing of, but it's usually within the year now.
It's down to 20 million, and it's usually within a year, maybe a year and a half at the most of production. I'm not doing multiples. The second thing is, you know, the embedded is basically based on what the projected volume of builds is. Tony's team has a real in-depth effort. We take the IHS data. We then narrow it down by OEM, geography, model, all of that. If there was a big shift of, you know, suddenly something happens in the world and IHS goes from 85 million down to 75 million, that would be an unpredicted shift. Everybody would be affected, like I'd be the last guy you'd worry about then, right? It'd be a lot of other people you'd worry more about then, right?
So we're pretty confident with that number, and we take a bit of conservative view of that. And then the connected is kind of the same way. It's okay, we know how many cars are produced, how many cars are gonna sell off the lots. That's usually, you know, there's only about maybe six months is the longest that a vehicle has that shift. So we're able to usually, like we may be off a little bit on a quarter, but probably within the year we're gonna be pretty good. Yeah. Without a lot of the noise, now I think that what our financials are pretty clean from a run rate standpoint of what we're doing. I think we have high visibility. We produce a backlog metric every other quarter. It's a roughly $1 billion five-year backlog.
So, you know, that backlog represents our contracts at our contracted price. Yes, there's a volume estimate that we have within there. But within the next 12 months, that volume is, you know, fairly visible. So I would say that, you know, the visibility standpoint, we're in like 90%. The only place we don't have a projection is here. So. Yeah. You know, we didn't have the Samsung forecast because it was supposed to go to court, and then you don't know how long that will take. These are due to go to court at the end of this calendar year, 2026. And Tony said we did it a little different. This one went through normal court. This one's going through the ITC, International Trade Commission.
So rather than sue for a dollar, I just sue to say you can't ship anymore to the U.S. And so it's a lot scarier for these guys, actually. But we have projected nothing for here because, so the way I look at it is there's only upside, here. So is there some chance there'll be an upside this year? Yeah, whatever you wanna gamble on, on these. But we don't worry too much about downsides unless there's macroeconomic shifts. Yes, sir. Are you saying growth will accelerate from here because your software is getting much better? Yeah. So there's two, well, there's probably more than two, but there's, I think of like two or three big things that will occur for growth. One is we said that growth of connected goes from 30%-40% to 90 %+ by the end of this decade.
We, that's additional revenue, right? That's new revenue that gets added on. That's revenue that from a stream that we don't have on that delta between 40 and 90 will be plus. The other thing is the technology. We produce a price per unit each quarter. We give you guys that, right? That's a backward-looking, over the last 12 months, average of what our price per unit is. It's grown from $45 to $50-something this year. The quotes that I have for our new technology, our PUI, are much higher than that. We haven't said publicly how much that is, but it's much higher. That will tell you that we'll get paid more per unit.
So, and then the third option is just getting more of our software stack as well because they, they could take different levels of our stack, right? So we're trying to sell them more of our software. We're trying to get paid more for our software, and we're trying to add connected. That's kind of the value stack of revenue growth. And that's only with automotive. And then you'd go outside of automotive is how we monetize the, the IP through, you know, through business development, similar to Samsung and other cases that we're seeing. Yes, sir. Probably the last question. Yeah. You mentioned that 90% of the cars are gonna be connected by 2030. Of the, that number of tokens that would be in a typical car like that, what percentage does Cerence have? Okay. So, I mean, you're asking me to project.
So if you take a look at it right now, the estimates are 40% of vehicles are connected today. 30% of what we shipped is connected. So I'm using kind of rounded numbers. So we think we have about 75% of the connected vehicles today. So if we continue at that percentage, then, you know, we'd be at, if it's 95%, I'd be at 70%-ish of the vehicles, would be connected probably for us. On the same slide, I think IHS had like 400 and something, maybe $500 of opportunity. Yes. That's for the OEM. I'm sorry? That's for the OEM. Oh, okay.
Yeah. It's not our revenue.
Yeah. I would love that. Yeah. Well, I thought that was per car. It is. Opportunity. That is what the OEM opportunity they believe is. Some of that's getting a renewal.
Some of that is adding additional features and capabilities on their end. One of the big, European OEMs is building their own charging infrastructure. So they're gonna, they ask us to not use the Google search, but actually go to their LLM and add their, if you're looking for charging locations, give their locations first, then the Google locations. And they're gonna get paid, you know, for that charging. And so they look at that as revenue. And so there's that, there's parking, restaurants, there's all kinds of things that they.
Just to be clear, it's not our revenue. Yeah. Yeah. Okay. Thank you. Yeah. Okay.
Thank you, guys. We're. Thank you. Appreciate it. Right on time.