Kayvon Tehranian: It's worked really well for us, and we are getting better at it, so we think we can do it even better and faster with LivePerson. So SoundHound has spent 20 plus years in technology innovation, and that's an area of strength for us. So we haven't had the need to go pay expensive dollars for technology acquisition or teams. That is our DNA. In enterprise AI, things take time. Adoption takes time. You have to develop a relationship with brands. You have to be part of their infrastructure. You have to go through a procurement. You start small even after you win the deal and then you scale with them. That takes time. And what we've learned is that those relationships and the time of integration and scale is just very important and very valuable. And that opened our eyes to the M&A strategy. And then as we were exploring that, we found amazing opportunities that I think are amazing for us as SoundHounds, not necessarily for a lot of other buyers. We find companies that have a great team, a great business, really strong customer relationships, and they are deeply integrated with their customers with a long history. But for some reasons, they are going through some stressful situations, and the combination really unlocks the value that was kind of trapped. And it really is the collaboration between the two teams that we come together and we give them what they need to really thrive. And it's really a collaboration between them. I wouldn't say that we are the only savior for their business. These are amazing teams. We collaborate together. And it's really amazing turnaround opportunities. And our first acquisition was just about two years ago. Same story, great team, great technology, great solution, great customer base, but they were declining. Within two years, they are on a continuous growth path. So we have completely turned them around with really strong growth. And we think we can do it live person. We can do it better and faster. Our last acquisition is just a few months ago, and we are already seeing signs of turnaround with them. And with live person, there are three... really important areas that we can address immediately. Financial stability, faster innovation, and faster modernization of their platform. And the story kind of gets fixed overnight. Of course, we have to execute after that. And I know you didn't ask, but if you exclude all of our acquisitions, Our core auto and IoT business was up 88% year-over-year this year. So we're doing really well organically with our core business, and we're really doing a good job turning around these businesses. And, you know, with the 52% growth year-over-year, that included some business components that are recent and have declined in them, and we did a great job and we're very optimistic about our outlook.
Gil: Got it. So just to kind of take that forward, asking the question, oh, this acquisition had X revenue a year ago, and now it has less than X revenue this year isn't really the right question because you're taking businesses that are either declining or in distress and you're being selective about what you're obtaining. So with that in mind, the live person, the public company and the expectations are for them to generate 200 million of revenue this year. And I believe your statement is that you're going to retain at least 100 of that, which is to say you do plan on some attrition of that business, planned attrition of that business moving forward before you start turning it around and growing again. Is that a correct understanding of your approach to the live person?
Mike Zagorski: Hey, Gil, I'll take this one. It's Mike Zagorski, Chief Operating Officer. I thought I'd jump in on the follow-up on that question. Just building on a lot of what Kayvon said, as Kayvon mentioned, this now will be our fifth acquisition. So we've built a bit of a foundation for integration. And ultimately, when we step back and look at each scenario, we try to account for all the variables. So it's a combination, as you can imagine, ARR, churn rates, pipeline. So in cases where there is a marked call it decline, we certainly factor that in. So we try to approach our path forward with the appropriate amount of conservatism. Ultimately, it's a matter of execution, right? Integrating the business, achieving new baseline of stability. And of course, we did announce our agentic platform, Oasis, two days ago. And of course, that is... That is a way truly to unify a lot of SoundHound's legacy business as well as the businesses of the companies we brought on board. So no longer does the technology stack become entrenched in the organization or even in the channel. It's a conversational AI platform where you build an agent once and it's omnichannel. And so that becomes part of our integration path and Again, as Tabon mentioned, we approach M&A with this best-of mindset, right? We combine what each company excels at. It's not one-sided. So we approach it humbly and conservatively, but with a goal for transformational turnarounds. So we've certainly started to see that. So in our goal with LivePerson, it's very much the same. And so ultimately, even though we are approaching it with the appropriate sort of long-term revenue targets in mind, our goal is to always exceed those expectations while making sure that we can hit what we're putting out there.
Gil: Got it. Makes sense. Thank you very much.
Kayvon Tehranian: Thank you.
Operator: Thank you. As a reminder, to ask a question, you'll need to press star 1-1 on your telephone and wait for your name to be announced. Our next question comes from the line of Mike Lattimore from Northland Capital Markets. Mike, your line is now open.
Vijay Devar: Hi, this is Vijay Devar for Mike Latimore. A couple of questions. What are the best near-term prospects for OASIS? Is it Amelia-based or the auto companies or the channel? Is it the new customers or established customers? Any comments on that, please?
Kayvon Tehranian: Yeah, this is Kayvon. I'll take this one. So we had, for example, a Fortune 100 insurance company that we had a renewal with them in Q1. It was an eight-digit size deal. They are getting great results from our agenting platform. Companies like that would be the first wave of beta users because it would make a very big impact to the quality and to the business, to their business, to our business. So there is a wave of upgrades and migration of the existing customers, starting with the larger ones, doing it very carefully because some of these integrations have been in place for years with thousands of incremental optimizations but all the new customers we expect will use Oasis going forward. And with Oasis, what used to take months now can be done in minutes because the concept behind Oasis is AI builds AI. So we used to have to go and take their documents, take their APIs, take their specs, and allocate resources and spend months to work with them to build. Now, we provide the vision to our AI and our AI builds the AI that can be deployed in multiple channels, in voice, in chat, in cars, and so on. Now, it's not just AI builds AI, it's also self-learning. So once it goes live, the data that comes in, all the interactions that the AI sees, it can improve itself, but it doesn't automatically improve itself because there's, you know, the concern of AI going in the wrong direction. It presents the improvements that it has designed for itself to a human operator for approval. And that is something that used to take constant maintenance with a large, you know, set of resources, now can be done automatically, again, with the human oversight to make sure we keep the AI in check. So going forward, we expect all of our customers to start using Oasis, which, again, will improve the quality, improve the speed of delivery and the profitability.
Vijay Devar: Got it. And how many Amelia customers might migrate to AgentDick AI this year?
Kayvon Tehranian: So maybe I'll step back a little bit because our new platform is Oasis. Oasis... combines all the great innovations of all the companies that have come together, you know, SoundHounds, Interactions, Amelia, and others. And, you know, if you look at the history of conversational AI for customer service, it was deterministic, then it was generative AI, then it was agentic, and now for us it's OASIS. and we have designed OASIS with all the great qualities that I mentioned, AI builds AI, self-learning, and so on. But we've also designed it to help us with integration of these various acquisitions because we don't want to kill the innovation that have taken place for 20 years of one acquisition to in favor of another platform, we want to actually inherit all the great qualities. And we designed the infrastructure of OASIS to be able to bring all of those innovations. So I'll give you an example. Our last acquisition was Interactions. They have a patented way for human oversight of AI. When AI knows it's not able to handle a question. Instead of transferring to a human, it asks a human to help it overcome that particular challenge. So we don't lose the containment completely because usually what other companies do is they just transfer to a human. So the AI loses its value. But the way they do it is they just Almost like, let me check with my supervisor. They go ask a human, what should I do in this case? And then they go back. The AI continues to handle the call. That's a very important innovation for a lot of their customers or other customers really valuing it. That innovation is coming into Oasis. So think of Oasis as a combination of the best of Amelia, the best of Sam, the best of interactions, the best of things we all combine, and hopefully the best of live person in the future. And we expect... Ultimately, all customers will migrate to OASIS. We are not sharing the exact number or the exact time frame, but that is the biggest priority for the company to migrate everyone to OASIS, migrate the teams to integrate to work on OASIS, and OASIS is going to be the foundation of our technology and solution going forward.
Vijay Devar: Got it. Thank you.
Operator: Thank you. Our next question comes from the line of Leo Carpio from Joseph Gunner. Leo, your line is now open.
Leo Carpio: Good afternoon, gentlemen. I just wanted to focus in terms of the competitive positioning. How are you thinking right now in terms of the competitive environment? Now you've got these large language model providers that are still focusing on native voice AI and identity capabilities. Are they still being a pressure, or are you seeing them, or are you thinking you've got a competitive mode at this point? And I've got a follow-up question.
Kayvon Tehranian: Yeah, so I will categorize two types of competitors. There's the big tech players and some of the frontier model providers, and then there is this newcomer's So I'll address both and how Sandhya is positioned against both of those. What you're seeing, first of all, what you're seeing is our customers, they don't want a vendor, they need a partner because the AI transformation, there's a mandate for AI transformation and they don't want an API, right? They want someone that can sit with them, listen to their pain points, address the pain points, help them dream big and help them achieve those dreams. And Samhan absolutely does that. We pitch ourselves as a partner for the transformation, not a vendor with some documentation. So that really eliminates some of the big tech players or the frontier model providers because that's really not their business to be a partner for all these hundreds and hundreds of enterprise customers. Sometimes they run science projects or some proof of concept, but we don't really see them as a head-to-head competitor. And then there are a number of companies that we call them Lego makers. They don't have their own technology. They are using an API for this, an API for that, a bunch of APIs and created solutions together. We win against them also in terms of quality because Sanhan has its own foundation model. And Sanhan brings the model from all the frontier models. So we promise to bring the best solution and best model to our customers no matter where it comes from. But for the majority of the cases, it's our own model. We have our own Polaris speech foundation model. It beats accuracy of all the other models we've tried. When we benchmark, it's 35% more accurate or more. When our customers test it, it's actually been better. Sometimes they report 80% more accurate. So because SoundHound has a DNA of the core technology itself and is a partner to our customers, we win on technology and we win on the partnership.
Mike Zagorski: I would just add, this is Mike again, just one more thing to add, because I wanted to tie it into a previous question on our M&A strategy. and how that actually makes us more competitive. And that's why the timing of this question is really good. Between M&A growth and customer acquisition and OASIS, what we're offering is scale, a proven level of scale. So we operate at a level of scale that is significant. It's global. It's enterprise. So customers have the confidence that we can deliver for them because we're doing it across industries and verticals. were production tested, meaning that we actually don't just work across the phone. We work in noisy environments. We work in vehicles. We are truly omnichannel. And that presents itself distribution opportunities. So if you envision an agentic future built on Oasis where it's a build one agent deployed anywhere, working with a company that can do it at scale, in environments, and across channels, omnichannel, that creates a very compelling environment package that isn't narrowed down to any one particular execution industry type for a pure play AI company to be able to do all those things is resonating with customers.
Leo Carpio: Okay. And then turning to the auto units in, have you been seeing any pricing pressure? I know auto company and automakers are trying to consolidate the AI vendors that they use. And just see if you're seeing that impact as you start gaining more share in that area.
Kayvon Tehranian: I think we are seeing the opposite because so there is the pre-gen AI solution, you know, automotive business for a number of years and they love our solution, both cloud and on the edge and lots, you know, content, all the vehicles, domains for car control and so on. And then And then we had the Gen AI upgrade, and that was an upgrade moment. So we basically gave our customers a choice of staying with the pre-Gen AI version, with the royalties that they were paying us, or upgrade, and they pay us more for it. And it's like one of those rare moments where, in the automotive industry, we were able to increase the revenue per unit. And then we kept... the force behind that. So we offered the live Gen AI after that. So we have the Gen AI version. And the live Gen AI is static information, but if people are asking for current news or something that happened today, we have a live Gen AI, just another upgrade that all of our customers are signing up for. And we are benefiting from increased revenue per unit And a lot of that actually becomes renewable. So instead of a one-time fee for the cloud, they pay for a certain duration and then there's a renewal after that. So that's another opportunity for growing revenue. Then the next milestone is agentic and OASIS. And that's going to bring even more capabilities, including voice commerce, which is something we've talked about for a while. And as I mentioned in my prepared remarks and in our first release today, we have a number of automakers and TV makers integrating our voice commerce, which is our agentic solution for devices, but it brings commerce opportunities. So while you're driving, you can order food. You can order coffee in the morning. You can order food for pickup on your way home. Ultimately, you can book parking or table reservation or other types of reservations. and that's a monetized opportunity that brings revenue for us, more revenue for us, more revenue for the car makers, more leads for the merchants, and that's the agentic upgrade that is also going to bring more revenue.
Leo Carpio: Okay, and I apologize. I just have one more question. Regarding LivePerson, this is going to be your fifth acquisition, and what's the biggest lesson that you learned from integrating Amelia and interactions that you're going to be applying to LivePerson?
Kayvon Tehranian: Oh, we've learned a thousand lessons. It's not just one thing. I think we keep getting better at it every time. You know, the first acquisition, I would say two years. After two years, we see a complete turnaround. It's on a path of growth. The revenue is growing. And a lot of complimentary. There are a lot of opportunities between, let's say, our first acquisition and our third acquisition. And the last one, which was... Not long ago, we are seeing the same signs within a year that we saw maybe in the first one in two years. And it's not just about one thing. Again, it's a thousand things. Integrating better, integrating faster, getting to know what they have, getting to know the team, meeting them. respecting all the innovation they have done, respecting every individual in the company and their vision, and getting in front of customers together. And in many cases, they become our leaders. We find amazing gems in these organizations with a lot of potential, and we put them in charge of different business units and different teams, and that has been working really well.
Gil: Okay, thank you.
Operator: Thank you. Our next question comes from the line of Scott Buck from Titan Partners. Scott, your line is now open.
Scott Buck: Hi, good afternoon, guys. Thanks for taking my questions. I just have one today. In terms of, you know, or as I think about past sustainable profitability, beyond scaling revenue, what are going to be the biggest drivers? Is it mix shift? Is it improved efficiencies, pricing power? How does that kind of break down over time?
Kayvon Tehranian: Yeah, so with our acquisitions, we get revenue opportunities. We also get cost synergy opportunities. And some of it is very obvious, like redundant cloud providers, for example, or redundant vendors. And we are constantly exploring those. In fact, we talked about it a couple of quarters ago. We took certain actions in Q1 for those cost reductions, and we expect to see the impact of that in the next one or two quarters. So it's a combination of exploring those cost synergies and revenue growth, but also making the right investments. So another thing I mentioned in my prepared remarks was we have a strategy of a very calculated, responsible investment in our foundation models. Historically, we've been very strong in speech foundation models, which we call Polaris. Polaris will be powering OASIS. But now we see an opportunity to enhance our foundation models to cover specialized language models and speech synthesis. Now, this is the right time for us to do this because of the scale that we have. As we migrate all these customers to OASIS, the cost of hitting frontier models increases. for example, can be substantial. So making this investment now will be a fraction of the cost that we would experience if we market all of our traffic to Oasis. And in a very short span of time, it will have a return on the investment. So our COGS will go down. and our costs will go down because of this investment that we're making today. And SoundHound will basically have the full stack of all the models to power the full experience of every interaction from our customer. And that type of innovation is going to reduce costs and improve profitability.
Scott Buck: Perfect. That's helpful. That's all I had, guys. I appreciate it. Thank you.
Kayvon Tehranian: Thank you, Scott.
Operator: Thank you. This concludes the question and answer session. Thank you for your participation in today's conference. This does conclude the program and you may now disconnect.
