Hello, and welcome to SoundHound Fourth Quarter 2024 Earnings Conference Call. At this time, all participants are in a listen-only mode. After the speakers' presentation, there will be a question-and-answer session. [Operator Instructions] I would now like to turn the conference over to Scott Smith. Sir, you may begin.

Good afternoon, and thank you for joining our fourth quarter 2024 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'll 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'd 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. We had our strongest quarter on record with $35 million in revenue, representing an increase of over 100% year-over-year. With that, we reached the top-end of our revenue guidance range for the full year, which we raised just last quarter. This speaks to the continued acceleration of our business. In fact, in the last five years, we have grown our top-line at a compound annual growth rate of more than 50%. And since going public less than three years ago, we have achieved significant growth in numerous key metrics; revenue increased by 8 times, queries increased by 7 times and bookings increased by 5 times. But the success we are achieving today didn't start three years ago. It began 20 years ago with a vision that is now becoming a reality. We made bold predictions early on and stayed committed to the path we envisioned. We built an enterprise-grade platform that powers some of the world's largest organizations. This laid the foundation for us to capitalize on the fast-growing voice AI market. Our ability to innovate at a rapid pace and maintain the agility of a pure-play high-growth disruptor allows us to deliver cutting-edge solutions with speed, accuracy and reliability. We are incredibly excited that we recently unveiled the third pillar of our business, our voice commerce ecosystem. SoundHound has pioneered this vision for years and the concept is simple yet powerful. The users of products powered by our voice assistant, our pillar one, can transact with businesses that are powered by our AI customer service offering, our pillar two. This forms the foundation of SoundHound's three-pillar business strategy and we expect it will increase our momentum further. At CES, we demonstrated the complete end-to-end experience of…

Thank you, Keyvan, and good afternoon, everyone. Q4 revenue was $34.5 million, increasing more than 100% year-over-year. For the full year, we grew by 85%, reaching the high-end of our guidance range with $85 million in revenue. Before we dive into the financials for the quarter, I'd like to reflect on 2024 and what we accomplished. Within the most dynamic and fastest-growing market in tech, we continue to execute our game plan, added to our portfolio of solutions and innovated with distinction in one of the most significant technological shifts in over a decade and quite possibly in our lifetime. 2024 was a breakthrough year for SoundHound as we diversified our business across products and industries, laying the foundation for scalable growth. The time for voice AI is now. Today, we work with 30% of the top quick-service restaurants and 70% of the top financial institutions in the world. We added new automotive customers, grew in healthcare, and this quarter, we entered a new high seven-figure deal with a large energy customer. We have significantly reduced our customer concentration. Our largest customer represented slightly more than 14% of revenue in 2024 versus comprising nearly half in 2023. We have transformed our business towards more predictable recurring revenue streams and have positioned ourselves for sustainable growth in 2025 and beyond. We continue to innovate with high velocity, further reinforcing our leadership in voice AI. The balance sheet is strong. We have capital flexibility to do the right things and our vision is now being realized. We finished the year with cumulative subscriptions and bookings backlog of nearly $1.2 billion, up over 75% year-over-year. As mentioned before, this metric is a measure of customer activity and gives current value to our existing contracts. The measure is based on contracts signed and gives…

Thank you. [Operator Instructions] Our first question comes from the line of Gil Luria with D.A. Davidson. Your line is open.

Yes, thank you. Good afternoon. Keyvan, there has been some pretty big breakthroughs recently in the efficiency of AI models. And those models are open-source and getting a lot smaller. I know you develop your own models, you use other models as well. Does the advances in this technology make it possible for you, especially in the restaurant drive-thru business to put more of the intelligence on the device and therefore, maybe reduce latency, make it more efficient and improve the performance? Is this something where you can leverage the recent advances?

Absolutely. In fact, we predicted this two years ago when we created our architecture to use LLMs to bring LLM into our conversational AI. We predicted that the models will become better and cheaper, and there will be multiple models that will be good at different things. Some of them will be made by SoundHound, some of them by third-party, some of them will be open-source. So, we created the architecture in a way that we can benefit from it as these advances are realized. So, those are absolutely good for us. The most immediate impact is that our platform becomes more accurate and the running cost goes down. But a lot of the -- our reliance on third-party APIs, we bring those in-house, including our own models or the open-source ones that we fine-tune. But as you said, also, we can bring this to the edge. We had a great demo at CES showing the entire large language model experience and generative AI without cloud connection.

Yeah. Lucky was very impressed when he was there. I'm sorry, I wasn't able to make it. Nitesh, for you that backlog number has grown quite a bit. Can you give us a couple more parameters around it? What's the duration of that backlog number? Maybe by vertical or by pillar, what are the different pieces in that? I think you said $1.2 billion.

Sure. Duration is pretty consistent with prior quarters, about six years. I think it's slightly north of six years. By vertical, I'll say, continue to have a good balance. If you go back a year-and-a-half ago, it was heavy on the automotive. Automotive continues to penetrate and grow, very excited coming out of CES and what we announced with voice commerce that we're getting great traction with a lot of deals. Keyvan mentioned some of those in the prepared remarks. Where we've seen the outpaced growth now for several quarters running has been in the restaurant space. And for us that the deals that we're signing continue to grow, it really continues to be about pace and scale and how we can ramp with them. And now, with Amelia in the mix and I mentioned some new customers and new industries that we're really excited about, but across predominantly healthcare, financial services, there's real strength. We had the new deal with the -- in the energy sector that we're excited about. We think there's a long runway that we can capitalize on there. So, what's wonderful about that metric, call it, this year as compared to when maybe we were talking about it last year is two-fold. Number one, the diversity point across industries, voice AI is agnostic. It can influence and benefit many different types of customers. The strength of it and then just the shift I mentioned in the prepared remarks around more recurring revenue basis in that metric, that's definitely that composition has grown. You see that a bit with our mix towards more pillar two revenue. So, I think all of that just continued strength and possibilities. And I also mentioned on the call that that's just the starting point for us. We continue to aggressively go after new deals that are not reflected in that metric. We are now having, particularly post some of these acquisitions, cross-sell and upsell opportunities, really exciting early days on. The -- our restaurant customers really leaning into some of the solutions that we got in the conversational AI with the Amelia acquisition or some of the Amelia customer base now leveraging some of our smart answering capabilities that we built-up on the SoundHound side. So, really, we can incubate a lot more with that existing customer base and certainly are continuing to grow and add more.

Got it. Thank you very much.

Thanks, Gil.

Thank you. Please standby for our next question. Our next question comes from the line of Thomas Blakey with Cantor. Your line is open.

Hey guys, thanks for taking my questions here. It's great to hear about all those POCs off of CES and probably even work you've done prior to that. Could we maybe click there and just talk about what the typical time to revenue is for POC? And I have a follow-up.

I'll start and Nitesh might add. So, with automotive, we can go -- we can run POCs very quickly, especially with the ones that we are live in the market. For example, last year, we ran several POCs with Stellantis, bringing generative AI and ChatGPT into the cars, and the result was incredible and they decided to actually go live with all their units with the SoundHound Chat AI that brings generative AI into it. So this time, we actually have a lot of new OEMs, as I mentioned, two Japanese OEMs and two German OEMs, several OEMs from US and Korea are going to run POCs. So, the timing of POCs is going to happen in the first half of this year for most of those. Sometimes we charge for those POCs, but the vision is so big in this case that we want to prove it as quickly as possible. And you asked about revenue impact. We actually don't have -- the third pillar monetization revenue is not -- is an upside this year. But the impact of the three-pillar vision is already being realized because our pillar one and pillar two customers are more eager to adopt our platform because they see that path to monetizable moments from the third pillar.

Yeah, maybe Tom, if I'll just add, I'm not sure if your question was specific to the POCs we were talking about coming out of CES, but maybe I'll -- if it's okay, I'll just expand your question more broadly how we see ramping of customer base that are slightly different depending on the industry and product vertical. So, within the automotive sector, traditionally, it's we get into a deal and we try to roll out and we get royalties on cars as they're shipped, we get the royalties. So, as we get more units out in market, that's what we've -- that's where we generate revenue. That's historically been the model. With restaurants, I think I've mentioned in prior calls, how really depends on the customer. One, where there's drive-thru, sometimes you have hardware requirements and we have to go through the process of making sure the drive-thrus have the proper microphones, headsets, display boards. And we've done a lot of innovation to really help our restaurant customers scale there. For example, we've innovated with like smaller footprint post with a display board that doesn't have the same permitting requirements and that allows us to go faster. So, there's a lot of experimentation and trial and working with great partners on the hardware side to provide that scale. On the phone ordering for restaurants, it's a little bit more of you can go much faster. We integrate with the point-of-sales and we can scale it technologically very quickly. Sometimes the gate in restaurants could be the corporate, sort of do they have franchisees that we need to have multiple conversations with, does corporate own all of the locations or does corporate have greater influence. So, those are some of the gating factors. And then, when we get into the enterprise business, where we're selling interactions and we can deploy the product pretty quickly. We have a professional services organization that specializes in getting the product ramped up. And then, really again kind of depends on the scale and interactions. And as we can get more and more use cases, we can move from customer-facing to internal employee-facing, that's where is where we scale the interaction base upon which we generate recurring revenue. So that -- sorry, to expand your question to something more broad, but there's different dimensions to all of the verticals.

Yeah. No, that was a great review, and I appreciate that more than you -- more than you know. And yeah, I was asking Keyvan a lot about the POCs from the pillar three, and it was great to see the flywheel in motion already impacting pillar one. It'd be exciting to see what that unit ramp looked like in pillar one in the coming year. Maybe in the follow-up, the seven-figure deal in the energy vertical, sounds like really exciting in terms of expanding into new verticals like this. First of all, from a administrative perspective, Nitesh, maybe any kind of like what's the economic impact in the current quarter or the early '25 that'd be helpful. And what does this kind of portend came on for? What SoundHound is bringing to outside the core auto restaurant and finserve and healthcare verticals? That'd be helpful. Thank you very much.

Sure, I'll hit the first part. So, it's a multi-year deal. What is great is the scale is really much larger than our traditional deals. So, in terms of interactions and the capabilities that we're bringing, it is a type of customer that pays early. So that's great. We've got -- actually there's -- from an economics and a cash perspective, there is accelerated cash collection. So that's the strength for us. But generally speaking, yeah, the bigger point of the enterprise business has given us great penetration into financial services where we talked about being in seven of the top 10 financial institutions. And then, in healthcare, where we've really got a large runway and we're investing because we see a lot of opportunity. And you can just imagine in people's personal interactions, you've got to set an appointment, sometimes you've got to get some results, you need to have a conversation with a nurse or a doctor, like all these interactions are very complicated through the traditional models and that's what we're modernizing and making easier for people. And now, in energy, whether you need to check is your power up or all these types of things, the use cases are tremendous and our platform scales very well. So, we're excited and, again, multi-year deals, so we can benefit from that relationship over the long-term and hope to expand even further.

Yeah. And it's a very repeatable offering, so we can go to other similar companies and offer them a more efficient way. And we are very proud and excited that we are expanding beyond restaurants. When we started our AI customer service initiatives number of years ago, we started with restaurants first and we said restaurants to us are like books first in Amazon. They started with books first. Now they sell everything. Let's start with restaurants, do it really well and then expand to other verticals. And now, we are well beyond restaurants. We are, as Nitesh said, we're in healthcare, financial services, government, military and retail, and we just added energy. And again, all of those are repeatable and getting faster -- adoption is getting faster.

Great to see the successes. Thank you for answering my questions.

You're welcome.

Thank you. Please standby for our next question. Our next question comes from the line of Mike Latimore with Northland Capital Markets. Your line is open.

Yeah, great. Thank you. On Amelia, they have both kind of customer engagement and then more internal IT support use cases. Are either one of those sort of more pronounced in the pipeline?

The customer service is more pronounced, but -- and that also is more aligned with SoundHound's core technology and history, but the IT automation is also very strategic because a lot of those customers can be customers of both and we can package and upsell and give them incentives to adopt both of them.

And there's probably -- just to give a couple of dimensions. So, one is they have the customer-facing, they also have employee-facing internal to enterprises that's an opportunity that's also there. And then, certainly the ITSM part. So, all of those are there. They also, I mentioned previously the mix, obviously -- we're investing and we value the SaaS-like software growth profile that they had, but they also have professional services that for us allows us to scale into enterprises and customize and implement. And then, they also have escalation support, which for us is the area I mentioned in the prepared remarks where we're looking at making sure we're -- all those contracts make sense on a long-term profitability standpoint. But what is wonderful with that and what we value tremendously is the data, and we get real-time production data that we can leverage, improve our models and then automate and scale that over time.

Got it. Good. And then just on the pricing or contracting model, is it most kind of your -- under that SaaS model where you get an annual subscription, or is there an element that's kind of usage-based?

Yeah, it's a bit of mix, but definitely interactions and there's different -- depending on the vertical and depending on the customer, there's either interaction-based, sometimes it's containment-based and success-based. I mean, generally speaking, obviously, we're creating AI outcomes to try to help deliver value for the customer. And a lot of times the customer is looking for, "Hey, I'll pay more for something that successfully contains a query." So, it is generally moving towards that SaaS-like subscription models, but there are sort of up until a certain volume, there's a certain number of interactions that are included. And then, above that, there's increased pricing.

Yeah. Okay. Thank you.

Thanks, Mike.

Thank you. Please standby for our next question. Our next question comes from the line of Scott Buck with H.C. Wainwright & Company. Your line is open.

Hey, good afternoon, guys. Thanks for taking my questions. I guess the first one, when I think about these secondary verticals, whether it's financial services, retail, telecom, healthcare, how are you prioritizing those? And do you have the capacity to go after all at once?

So, I will start. Build up, we are from the origins of voice AI company. And so, Keyvan in particular, but the founding team and all the engineers that grew with the company have built an amazing engine and a platform that we're now able to scale across multiple verticals. So, we're a platform technology company. And then, ultimately, it comes down to where do we want to practice? Where is the biggest use case and benefit to a customer? So, we knew early on, automotive there's real benefits to being able to communicate with your car while driving for safety reasons and so forth, and there was investment from the OEMs to try to scale into that. As Keyvan mentioned, we went into customer service, we started with restaurants, and that one made a ton of sense because our technology differentiates based on how good it is, how accurate it is, and when you order food, you don't want just any old pizza, you want the pepperoni and sausage specifically, and that's what our technology and product does really well at. So, it made a lot of sense to focus on that. As we expanded into enterprise verticals and brought some of our -- some of these new capabilities with these acquisitions, we see a lot of runway. And I mentioned this, I think after the acquisition, it's not like we were -- we had a lot of runway with restaurants. I think I mentioned the scale of opportunity in the prepared remarks. We weren't maybe necessarily rushing to jump into the other verticals, but the opportunity with a really unique asset and a great time made sense for us. And I think really where the investment point is then to -- Scott, to your question is, from a…

Great. I appreciate all that. And then, I want to follow up on the earlier question about cross-selling and upselling. It sounds like you're already starting to see some momentum there. I'm curious, where is the sweet spot? Is that nine months out, 18 months out? I mean, what does that look like, I guess, the ramp or the cadence of how that's executed?

It really depends. I wish -- I mean, I wish I'd give you a precise answer, but we're already in a -- I know of a few really, really interesting ones that can be very meaningful to us and they're just not far enough along for me to give you a sharp timeline. But there are people who are with some other players out there in a renewal period that's coming up this year and we want to pounce on that opportunity and say we can take -- get a competitive win. So, there are deals in motion as we speak with either we've got to wait for sort of a renewal timeline or in some cases, is there a way to kind of breakthrough. I see near-term opportunities, but certainly, as you get into the six, nine months and into next year, we think there's a lot of opportunities. Part of the upsell is also integrating product stacks. We have a smart answering solution. We're really excited about it and it's an advanced answering capability that's rolling out we mentioned it in like fitness centers around the US and it's a solution that helps with upselling membership and so forth. And Amelia had a similar product and now bringing those kind of together and integrating and allowing us to scale across enterprise and into some of the retail that we were already going after. There's a lot of things to go after. And there's -- and so, I hesitate to put a timeline on it. I could just tell you we're aggressively going after it. And maybe one last point is architecturally things just -- we're still pretty early in the integration plan. Things have been going well. But just to make things work on the ground, sometimes you're really getting aligned comp plans and making sure that people are incented properly. I will tell you, generically, we really want to move fast on capturing customer opportunities. So, if we marginally air on the side of double paying here or there between two groups like that's okay in the near-term, we obviously need to normalize that over the medium-term. But really for us, it's about capturing scale. That's what we've been focused on for and with the -- just in recent past and we'll continue to going forward.

Perfect. I appreciate all the added detail, and thanks for taking my questions, guys.

Thank you.

Thank you. [Operator Instructions] Please standby for our next question. Our next question comes from the line of Glenn Mattson with Ladenburg. Your line is open.

Hi, thanks for taking the question. Keyvan, curious, you mentioned military a couple of times. Maybe wondering if you could expand if you're able to on what you're doing there? And just between the military and the large energy deal, just curious how that kind of these new verticals fall into the sales funnel, or do you have guys out there looking at new verticals that are -- haven't historically been the sweet spot?

Yeah. So, it's all similar at the core conversational AI and voice enabling the world and there are a lot of opportunities for making things safer and more efficient. And so, the one we announced was we have a partner that helps us bring our solution to the government and military. Along the same lines, we announced a partnership with a company called Rekor that will bring voice AI to police cars and fire trucks and so on. And I said this recently also that we are -- we benefit when the economy is good people want to invest in AI and innovation and that brings them to us. And then, when economy is challenging, people want to invest in automation and cost-saving, that also brings them to us. So, you will see both types of deals. I mean, there are transactional deals. So, in restaurants, we generate new lease, increase revenue, we upsell, and there is some cost-saving also, but there are -- when we automate calls and interactions that are very routine, that has the cost-saving elements.

Great. Thanks, Keyvan. And just curious on just with the guidance and -- just curious, I know it's a lot more SaaS than it used to be. There used to be some lumpiness a couple of years back when you get a large kind of on-prem auto deal. Has that gone away? Is that still something we should think about? And just kind of overall your -- none of the business has shifted quite dramatically with the addition of Amelia, and these other things, just your ability to how confident you feel about the general forecast and all that?

Yeah. With the greater mix of SaaS and it certainly helps with predictability. And we -- with respect to sort of historical deals, we -- they're still out there, I should say. Like we can -- sometimes in customer conversations when we're deploying particularly our edge product, if you know a customer wants -- we want commitment because we're going to invest in a product like we want to make sure that the volume is there to provide a sufficient return. And then, for a commitment, sometimes the customer wants to negotiate clearly on price. Like those are discussions that are ongoing all the time. So, the benefit of having a more diversified revenue base is we have a lot more leverage into those discussions and we can kind of go, well, here's sort of our clearing price. And if it doesn't sort of meet the threshold, then we're willing to either pause or just say, let's move on to some other type of discussion. But I do think general mix is certainly much more SaaS and recurring. I mentioned with the acquisition, there are different pieces of what we got with the Amelia business. There's a professional services business that is really important for us in terms of activating implementation and customization. Depending on the customer environment, the scale and complexity on those deals can be different. So that can create a little bit of noise quarter-to-quarter. What we'd love to do is obviously take a standard implementation package or customization package, and then make it much more efficient going forward, so we could be faster to scale. And there's also escalation support, which generally can be recurring, but we're looking at those contracts in great detail and going exactly which one makes sense for us over the long…

Great. Very helpful. Thanks very much.

Thank you.

Thanks, Glenn.

Please standby for our next question. Our next question comes from the line of Leo Carpio with Joseph Gunnar. Your line is open.

Good afternoon, gentlemen. I've got two questions. The first question is regarding your platform. When you look at the platform, are there any missing technologies or systems that you think you still need that you need to go and acquire? And then, secondly, and on the auto POC contracts, how quickly can those POCs move from testing to full contracts and revenue? Could it be like a 12- to 18-month horizon? Thank you.

Yes. So, on your first question, we try to be a complete like one-stop shop for our customers and provide them everything they need. In the majority of those cases, those technologies are built in-house over the last 20 years and we are very proud to outperform a lot of our competitors. For example, I mentioned our latest Polaris speech recognition model beats Google by more than 20% and others, but we are also not shy about partnerships. For example, if a partner has a better text-to-speech that sounds better and some of our customers prefer that, we give them that choice. So, we kind of don't have its own text-to-speech, but it also we have partnerships that allow us to offer third-party text-to-speech to our customers. And that I think is a source of strength for SoundHound to not be shy about partnerships in addition to building in-house. So, ultimately, for our customers, they work with us. Sometimes we source the technologies from third parties, but ultimately they get it all from us. And then, you -- what was your second question? I'm sorry.

The second question was...

Yes, the POC. If they are already live with us in production is -- in most cases, it's just a flip of a switch. So, we take a number of units based on their unique IDs and we enable the new feature in those. And we can do that very quickly. We can run the pilots and do the audit and get the results and then to go live with more units is [indiscernible]. So, that's how it worked when we brought generative AI and large language models into Stellantis vehicles in Europe last year. If it's a new OEM that is not using our voice assistant by default, it would have to be -- it's a longer process. But that's also the incredible power of our third pillar, it is bringing a lot more OEMs to us to choose us for their default assistance going forward because of the monetization opportunity.

All right. Thank you.

Thank you. Ladies and gentlemen, I'm showing no further questions in the queue. That concludes today's conference call. Thank you for your participation. You may now disconnect.