Good day, everyone, and thank you for standing by. Welcome to the SoundHound Q4 2025 Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded. I would now like to hand the conference over to your first speaker today, Scott Smith, Head of Investor Relations. Please go ahead.

Good afternoon, and thank you for joining our fourth quarter and full year 2025 conference call. With me today is our CEO, Keyvan Mohajer; and our CFO, Nitesh Sharan. We will begin with some short remarks before moving to Q&A. We'd also like to remind everyone that we will be making forward-looking statements on this call. Actual results could differ materially from those suggested by our forward-looking statements. Please refer to our filings with the SEC for a detailed discussion of the risks and uncertainties that could affect our business and for discussion statements that qualify as forward-looking statements. In addition, we may discuss certain non-GAAP measures. Please refer to today's press release for more detailed financial results and further details on the definitions, limitations, and uses of those measures and reconciliations from GAAP to non-GAAP. Also note that the forward-looking statements on this call are based on information available to us as of today's date. We undertake no obligation to update any forward-looking statements, except as required by law. Finally, this call is being audio webcast in its entirety on our Investor Relations website. An audio replay will be available following today's call. With that, I would like to turn the call over to our CEO, Keyvan Mohajer. Please go ahead, Keyvan.

Thank you, Scott, and thank you to everyone for joining the call today. 2025 was a record year for SoundHound, nearly doubling our revenue year-over-year. We also had a record fourth quarter. Revenue was up 59%, while all key profit metrics improved. We broke another record in Q4. We signed over 100 customer deals, making it our biggest quarter yet. We won across different industries in a variety of regions. Just to name a few, we signed a new prominent automotive logo in Japan to use our AI assistant with a 7-digit unit commitment. In the U.S., we signed a multiyear deal with one of the largest telecommunications companies in the world to use our technology. We signed a multiyear global deal with one of the largest athletic shoes and apparel companies to power their AI customer service. We closed deals with health care providers, universities, insurance companies, financial institutions, e-commerce merchants, retail, military, and many more. Our execution with channel partners was also exceptional with multiple 7-figure deals in 2025. I'll dive into other business highlights specific to Q4 shortly. But first, I wanted to touch on a few recent market dynamics. The power of AI is disrupting traditional software and services companies, and this is creating further tailwinds for SoundHound. In this inevitable AI transformation, companies need a partner like SoundHound to help them rapidly reinvent themselves. We partner with our customers to overcome their challenges and achieve their ambitions, creating incredible end user experiences for their employees and customers. With the exponential advances in AI, we believe we are entering a new era where companies with deep tech and data moats will create the most value. This makes SoundHound very well positioned with decades of deep tech innovation and data accumulation. SoundHound AI was founded with a…

Thank you, Keyvan, and good afternoon, everyone. Q4 was our strongest quarter with $55.1 million in revenue, up 59% and improvements across all profit measures. For the full year, we delivered $169 million in revenue, up 99% versus the prior year, and up more than fivefold in the few years that we have been a public company. We achieved this record performance through our disruptive technology, breakthrough innovation, hyperresponsiveness to customers and by scaling across our broadening enterprise portfolio. And we operationalize this with cost discipline, driving a clear pathway to breakeven profitability. The market momentum in our space continues to accelerate. Generative AI, Agentic AI, and Voice AI are now base level customer requirements. Customer service is undergoing a once-in-a-generation disruption and enterprises are clamoring for innovators like us to provide high customer engagement solutions to improve their top and bottom lines. From the beginning, we have built our business to deliver successful AI-driven outcomes and our pricing architecture is purpose-built for that. In a world where seat-based pricing models are quickly becoming antiquated because of their deteriorating price/value equations, our Agentic solutions seamlessly drive outcome-focused consumption and success rates that create economic incentives fully aligned with our customers. That's a sustainable model. It's a differentiated moat with our entrenchment deepening. Let me share some examples across our business. We have been growing the automotive installed base for years, and our monthly active users continue to expand rapidly with Q4 growth in excess of 50% year-on-year. More notably, their query activity or usage continues to accelerate with Q4 audio queries up roughly 75% from the prior year. And note that this is only cloud-based queries. We also offer edge-based solutions that don't require internet connectivity, so these volume metrics meaningfully understate the full auto customer engagement. The volume of…

[Operator Instructions] Your first question comes from the line of Scott Buck from H.C. Wainwright & Company.

As we went through the 4Q highlights, clearly, a lot of balls in the air. I'm curious, how are you handling from a deployment and customer service capacity standpoint? Are you starting to feel a little constrained?

Thanks for the question. We are definitely doing a lot. And I've been saying for the past few quarters that this is the time for us to do more, partly because, yes, we are in many industries, but the ingredients we are using to power these experiences is the same. So the decades of work we've done to build the best-in-class speech recognition, conversational AI, Agentic orchestration, all of that is the same regardless of whether we are in automotive or we are providing customer service for a health care company or insurance company. And we are -- because of the advances in AI, we are actually able to deploy faster, go live faster, develop faster. And we are able to keep up with the demand with fewer people and less resources. So demand is going up and what we need to do to deliver to these customers, the resources requirement is actually going down. So we expect that to be a further tailwind for us.

Great. That's helpful. And then I wanted to ask, you called out a number of renewals. Can you talk a little bit about any changes in pricing or upselling you're seeing as you go through the renewal process with customers?

Yes. So in -- we have customers that we've had for a while, for example, some of the automotive logos that we've had for a number of years. These renewals are actually an upsell moment because we bring them the Agentic solution. The Gen AI solution that we developed 3 years ago, that was an upsell moment. Now we have the Agentic solution. That's another upsell moment. So it's basically the renewal with a price increase and sometimes with a bigger volume commitment. And we are seeing the same in customer service, again, especially with the customers we've had for a number of years, the Agentic platform that is an upgrade, partly, it could come at a higher price. And for deals that are based on containment rate, we expect to generate more revenue because we contain more of the incoming calls. For example, if I'm going to use some numbers, but in the industry, there's no like one number for containment rate because it depends on the use case. We've seen, for example, in the use case, a 30% containment go to 70%, 80%, sometimes over 90%. That means we handle more than 90% of the incoming calls without kicking to a human, and we get paid more as we contain more calls. So some of the increase in revenue is just going to come from upgrading our existing customers to the Agentic solution even without a renewal and even without increasing the price, it's going to generate more revenue for us.

Your next question comes from the line of Brian Schwartz with Oppenheimer.

Congratulations on a very good year. Keyvan, I want to start with you. And your enterprise AI business clearly has strong momentum, especially in the higher regulated industries that you pointed out. You're building deeper entrenchment. But in the market, certainly over the last like 3, 4, 5 months, there's been a lot of fear about software companies' long-term growth that these larger LLM providers are going to be able to just build workflows above software companies' platforms and bypass them and it's going to be much more challenging for companies to grow. So I was hoping you could address that, how you see the durability of the enterprise AI business as we enter this agentic era? And then I have a follow-up for Nitesh.

Sure. That's a great question. So I think there are 2 flavors of that question. One is what's happening to software and services companies, so not necessarily SoundHound. And that has been a tailwind for us for the past 3-plus years because of Gen AI. So the concept of that is the automation is coming, and that is going to disrupt services companies and SaaS companies. Everything is going to get automated and more and more automated. And that's not a new thing. It's been -- we've been benefiting from that for 3 years, and these companies that want to automate basically come to SoundHound to help them automate. There is maybe a second flavor of that question, which is what happens to companies like SoundHound with the latest advances in AI where software development is becoming easier. We think of that as another tailwind. And a really good analogy that has resonated with me is imagine the internet companies when we had connections with dial-up modems. So good internet companies were there, and they were delivering their websites to their visitors. And then dial-up modem became broadband internet, so bandwidths went up in orders of magnitude. That was a great thing for internet companies. Now some didn't maybe reinvent themselves quickly, but the services that these internet companies could provide became a lot more richer and powerful. And we feel the same way that we can move faster, like the earlier question that was asked like can we -- we have so many customers, how quickly can we deliver? We use AI to build AI to deliver to the customers that want AI. And we think that pace is going to be better for us. The quality is going to be higher because we are utilizing these advances in AI.

I just wanted to ask you that question, Keyvan, because you're such a pioneer in terms of technology in this industry. I appreciate you sharing your perspective. The question I have for Nitesh is just thinking about how you're thinking about planning the progression of the efficiency of the business. So maybe asking it in terms of the operating profile here in 2026. Clearly, the business is accelerating. I think you're gaining efficiency in your development from your -- as Keyvan talked about with your own Agentic and AI. But how do you think about the rest of the investment profile? Are you looking to accelerate your investments? Are you looking to keep your margins stable where they are today? Are you looking to show improvement in terms of the efficiency and the EBITDA margin 2026?

Thanks, Brian. I'll take that from a couple of different angles. First, with respect to continuing that AI efficiency play, I think there's a multitude of ways as that plays out here. #1, our efficiency in product development is better. Certainly, we're seeing that. I think our efficiency in deployment and delivery is better. We're seeing that. And then even just operationally, we're all across the company, utilizing tools that may not be core to what SoundHound develops. But certainly, in my G&A function, there's definitely a number of areas that we're driving efficiencies. And I think that's a responsibility that all of us take to leverage the latest and greatest to be responsible with our costs. #2, to your question on just the profile, I'll go back to my prepared remarks a little bit. We are now shifting from an era, I'd say, from the origins of SoundHound where we were heavily investing primarily in our innovation, but maybe more recently in building up the go-to-market capabilities to now this area -- to this era of breakeven. And we're not trying to be super precise one quarter delivering one specific number. But more importantly, we start with the premise that we are in the very early innings of massive transformational shift, whether that's the LLM-driven capabilities that we're seeing with these amazing engines to the voice AI era of how we're able to engage with consumers and customers in really unique, more efficient, seamless ways to get all sorts of transactions done or obviously now to Agentic and how we can deliver great platforms and capabilities to enhance customer capabilities and solutions, predominantly, I think, for us in the customer service space and rewrite how those traditional sort of ways that people get their billing inquiries resolved or they…

Your next question comes from the line of Gil Luria with D.A. Davidson.

Great. This is [ Lucky ] on for Gil Luria. You guys had pretty strong traction, it sounds like with auto OEMs, especially on net new in the quarter despite previous headwinds from tariffs impacting the industry. I guess is there anything to call out as to why you had particularly prominent success in that vertical this quarter?

Yes. It actually was a great year in automotive for us. Earlier in the year, we closed with a big Chinese OEM, also with multiple millions of units committed. Then we had a robot maker in China, then we had more deals in India. And then we're very proud that we have won a prominent logo in Japan. And I think it's because of the great solutions we've created. We've been in automotive for a number of years. We are well known for having the best solution, and we partner with our customers to achieve their vision and ambition. Our Pillar 3 vision is paying off. It's -- we predicted a flywheel effect from Pillar 3. And to summarize it quickly, we power cars and TVs and devices in Pillar 1 and then we power customer service for merchants in Pillar 2, and then we connect them in Pillar 3 together. So while you're driving, you can order a copy, you can book appointments, you can reserve tables. And that's a monetizable moment. We call it the voice commerce. And just the concept of being able to deliver value to the drivers while generating revenue for us and share that revenue with the OEM is creating a flywheel for us and a lot of OEMs are choosing to work with us. So it's a combination of being a great partner, having the best technology and the concept of monetization with our Agentic AI in the cars.

And hopefully, Lucky, you're also noticing that we're seeing this growth in the fastest-growing markets, too. So oftentimes in the fastest-growing markets, it's sort of where they want the best-of-breed technology, and I think that's what's playing out here as well.

I think that makes a lot of sense. Maybe the last question from me. As you enter your next phase of growth here, you touched on it already, but can you kind of stack rank the top investment priorities to capture the opportunity in front of you? And any update on your M&A strategy in light of the broad decline in valuations across software here recently?

Maybe I'll take some of it and Nitesh can add. Our Agentic platform is an area we are heavily investing in. It delivers a much better user experience. It has a higher containment rate. And then the latest version of our platform, it uses AI to create AI. So a lot of the experiences that it used to take us maybe a team of -- a large team of developers, weeks and months to deliver. You just tell it what you need to do and it just does it for you. So that is going to allow us to move faster, deliver better quality, have a higher containment rates, win more customers. So that's one area of focus for us. And that is going to be everywhere. It's going to be in automotive. It's going to be in customer service, Pillar 1, Pillar 2, Pillar 3. The next area that I would highlight is voice commerce. We've been -- it's a vision that we pioneered. We are ahead of others in the space. It's a great idea, but it will take time for people to catch up because we are -- we have the largest number of merchants that are using our voice AI, like we are the largest number of restaurants, and we have a huge footprint in cars and TVs and devices. So we are in a very good position to bring this to market.

Yes. And I can add on the M&A part of your question. I think a couple of years in, I'll just say, I think the M&A approach we're taking seems right to us. I think we'll keep being mindful about what's in the marketplace and potential partnerships being aware of opportunities to combine. And so far, they've been companies that have had really amazing customer relationships we've been able to harness together and bring our innovation jointly to expand those relationships, use some of those landing points to expand broader in the relevant industries. And I think we'll continue to seek some of those opportunities. The -- because we've done a few, I'd say there's a lot more inbound also interest. And so we have a pretty strict and disciplined formula or methodology we go through in assessing if one may make sense, and we'll just keep that discipline. There's plenty that we look at that don't make sense for us. And I think we put a lot of scrutiny into when it might. Again, since things are moving really fast, we've said before, and I'll just repeat that we know we can do a lot of good things with what we've built here, but we don't want to be insular. And so we want to be aware of all the great things and partnerships that we can build. So I think M&A, certainly, thinking out the next few years, will continue to be a really important part of our muscle.

Your next question comes from the line of Mike Latimore with Northland Capital Markets.

This is Vijay Devar for Mike Latimore. A couple of questions. So one, how many Amelia customers are live on your Agentic AI version 7.3 and are likely to go live this year?

Yes. Hello, Vijay. We said last time, and I think we're just continuing to make progress that we had early last year sort of early adopter program where we piloted with, I think, at the time, we said 15 or so customers and that ramped through the summer. And we've been on this pathway of migration where we assume that we kind of are on a pathway to get the majority -- vast majority, I think, over 75% migrated over by the middle of this year. So we're continuing to be on that pathway, making incremental progress every quarter that we go by, and we continue to see that in Q4, and we're already continuing to see acceleration in Q1. One of the good things with our new Agentic platform that we highlighted at the Consumer Electronics Show that we talked about in the prepared remarks is that's something that we're getting a lot of early traction from customers on the channel our Head of Sales there and saying that the customer feedback has been superb. And so one of the great things we've been trying to deploy is sort of automated migration paths to allow people to migrate from prior version to Amelia 7.3 onto the new version that we're rolling out. So we are excited that our migration path in, especially going back to one of the earlier questions about the efficiency now with delivery and AI and what it's allowing us, it is allowing for more rapid migration patterns.

Got it. So when the customer moves to 7.3, is there incremental revenue to SoundHound?

Yes. So as I mentioned, just a pure higher containment rate is expected to increase our revenue. Many of our deals, we get paid when we successfully avoid a caller to go to a human. And by going to Agentic, we've seen just as an example, numbers going from like a 30% containment to over 90% containment. In some cases, we also get paid more for the upgrading to Agentic. So it's a mix of both, but either way directionally positive for revenue.

The other thing we've seen with the recent versions is sort of interactions that previously fell outside or would have to get escalated or things we can capture now at a much greater rate. So all of that is incremental revenue for us.

[Operator Instructions] Your next question comes from the line of James Fish with Piper Sandler.

Just on the CX side of things, how is Amelia effectively winning new customers versus the contact center pure play, the CRM offerings, and even some of the other stand-alone AI solutions out there? Really, what's making them different that's resonating with customers? And then I've got a follow-up.

Yes. So we -- so first of all, SoundHound, with the acquisitions we've made, we are a combination of teams and companies with several decades of collective experience in customer service. So we've been at this for a long time. And we have deep technology and we have data. And we are also very deep in a lot of these industries like health care, insurance, banking and so on and the reputation, the experience, the relationships all help us. But if you zoom out, and this might answer an earlier question that maybe I didn't quite answer from Scott was, if a customer chooses to work with a big tech, that's a very high-risk decision because when you choose a company like Google or OpenAI as your vendor and by the way, these companies are not really deep into customer service. They provide tools for others like SoundHound. But if they just go deep with one big tech, they're going to miss out on innovation that might come out of other big tech. So if you -- for example, you're betting on Google models and then if OpenAI creates an ecosystem that benefits the world, you might miss out on it or vice versa. So -- but SoundHound provides -- our philosophy is we provide the best technology, the best model to our customers, no matter where it comes from. Most of the time it comes from us because we have our own models that we've created over a long period of time with a lot of data and we compare it with the big tech models and we beat them in accuracy, speed and cost. But if for some edge cases or use cases or scenarios, a big tech model is better, we bring it to our platform and…

And maybe I can add one thing, Jim. And I said this in the prepared remarks, but it's salient to your question. I think fundamentally, legacy -- this goes back to the earlier question on, I think, for Brian on disruption in software. Models that were built seat-based pricing that value was on let's get more users using our tool, and it wasn't as directly tied to customer outcomes, that is at risk because the tools are getting so good and we're building those. And our solutions are directly -- our economic model, our pricing model is directly tied to customers achieving or seeing real value. Did a prescription get refilled? Did appointment get booked? Did food get ordered? So I think just architecturally, we're better to Keyvan's point. And I'd say also the economic incentive alignment is also an advantage for us.

Got it. Very detailed answer, guys. Appreciate that. Maybe, Nitesh, for you. Just a -- I'm getting it asked here after hours, just to hammer home a fine point. Is there much further M&A included in the annual guide? Obviously, you guys have a ton of opportunity. You guys have a strategic sense of doing acquisitions and folding them in. Just trying to understand if there's further M&A contemplated in the guide. And given the environment, there's a lot of sensitivity to stock-based comp. You guys are a bit of an outlier here. I guess, how are you handling your stock comp going forward and the dilution we've seen historically?

Sure. Thanks, Jim. For the first part of that, to be very direct, no, our guidance does not contemplate M&A that we haven't done or that is not baked into the outlook. Our outlook reflects the base of business, the pipeline of deals we have in motion, the existing customer activity that's recurring and we're upselling and expanding, and that's what's reflected in the outlook range. Now as I mentioned in an earlier call, certainly, there are M&A ideas out there and conversations we have from time to time and inbounds we receive that if they happen, we will -- if they're meaningful and material and we need to update an outlook, we will certainly do that like we have in the past. Your second question on stock-based comp, yes. So I'll start with this general point at SoundHound, which I think a lot of us internally have a lot of pride around, and I give all credit to Keyvan and the culture he's built. Like this is a company where we want all employees to participate in the full contribution. And so we do distribute equity to our entire company. And I don't think that's the same as every other company. So we feel pride because everybody here is an owner. And we've also seen historically a lot of volatility that makes in some of the math on our P&L, especially with the acquisitions, you get mark-to-market activity that when you have a stock that's so volatile, you'll distribute a grant. By the time it gets valued for P&L purposes, it's at $20 and then that amortizes over 4 years. So that is something that's a little different because we're such a high-vol stock. But I take the general element of your question. We do think of the economic impact certainly of dilution and certainly of where stock comp fits in our overall compensation profile. And we want to be competitive. We want to attract all the right talent and make sure we have the right people. To an earlier question, though, we are always vigilantly looking at our cost structure and can we do things more efficiently. And we have historically taken cost actions, including to our people when we needed to. But -- so I'm mindful of the element of your question. I think we do try to be equitable in our distribution of stock comp. We are mindful of the dilution impact. We are mindful to your point on it is probably a higher percent of revenue than you might see elsewhere. I think as we scale, you will see that normalize certainly as you will with some of the other operating lines. But again, I'll close with what I started with, which is I think we generally do feel a sense of pride that we do distribute equity to all -- everybody at this company, and that makes us all combined owners in the outcome and success of what we build.

Thank you. I'm showing no further questions at this time. Thank you for your participation in today's conference. This does conclude the program, and you may now disconnect.