Operator:
Good morning and good evening, ladies and gentlemen. Thank you for standing by and welcome to Tuya Inc.’s First Quarter 2025 Earnings Conference Call. [Operator Instructions] Please be advised that today’s conference is being recorded. I’ll now turn the call over to your first speaker today, Mr. Reg Chai, Investor Relations Director of Tuya. Please go ahead, Reg.

Reg Chai:
Thank you. Hello, everyone. Welcome to our first quarter 2025 earnings call. Joining us today are Founder and CEO of Tuya, Mr. Jerry Wang; and our Co-Founder and CFO, Mr. Alex Yang. The first quarter 2025 financial results and webcast of this conference call are available at ir.tuya.com. A replay of this call will also be available on our IR website in a few hours. Before we continue, I refer you to our Safe Harbor statement in our earnings press release, which applies to this call as we will make forward-looking statements. With that, I will now turn the call to our Founder and CEO, Mr. Jerry Wang. Jerry will deliver his remarks in Chinese, which will be followed by corresponding English translation. Jerry, please.

Jerry Wang:
[Foreign Language] Hello, everyone. Thank you for joining Tuya's earnings call for the first quarter of 2025. [Foreign Language] This quarter, building in a high gross base from Q1 last year, we achieved approximately 21% year-over-year revenue growth while maintaining a solid gross margin. Our stable team structure and organizational management supported a strong operating leverage coupled with the continued declining historical share-based compensation expenses, which achieved a record GAAP net profit in what is traditionally an offseason. With the net profit margin of around 15%, a significant improvement. [Foreign Language] Since the beginning of the year, the micro environment has been like a roller coaster from the rapid advancement of AI technology to trade fluctuations under geopolitical pressures. The smart consumer electronics sector and its upstream and downstream supply chains have faced immense challenges. Some of these we can manage through internal adjustments, while others require an industrial-wide response. In such a context, it's even more critical for us to stay the course strategically and leverage our unique value to navigate uncertainties and embrace a new landscape. [Foreign Language] Following the launch of our Tuya AI agent development platform last year, we continued to expand our AI capabilities from cloud to device in Q1, covering a wide range of AI agents and smart solutions. On April 23, we hosted our first Global Developer Conference of 2025 unveiling for AIoT development engines and a series of AI hardware solutions to support global developers and partners in commercializing AI powered products. The conference attracted over 2700 on-site attendees, including enthusiastic participations from customers across Europe, Latin America, South Africa, and Asia Pacific. We firmly believe that Tuya's unique platform model not only facilitates the deep integration of AI and smart hardware, but also drives continuous improvement in industrial-wide intelligence adoption. [Foreign Language] Operationally, we will maintain efficient organization -- execution safeguard a favorable environment for R&D and technology development amid external uncertainties, and remain discipline in managing costs to maximize operating profits and shareholder returns. [Foreign Language] Now, I'll turn it over to Tuya's Co-founder and CFO, Alex Yang, who will walk you through the financial results and business highlights.

Alex Yang:
Hello, everyone. This is Alex. Please note that all the figures mentioned below are in US dollars and all the comparisons are year-over-year based. In Q1 of 2025, we delivered approximately $74.7 million revenue and representing a year-over-year growth of about 21.1%. The revenue grows by roughly 70.9% with strong performance across major categories led by home appliances and followed by security censoring and electrical analyzing products. SaaS and others generated about $10 million revenue, growing approximately 15.5% year-over- year, primarily driven by the steady growth in SaaS value-added services. Smart solution revenue reaches approximately $11 million with year-over-year increase about 47.1% with excellent growth in smart video products, central control, innovative appliances, and professional lighting solutions. From a regional revenue source perspective, Europe accounted for about one-third of the total revenue. And then comes after that is Asia Pacific, excluding China, and then China and Latin America -- Latin America, each of them contributing around 15%. And then with many regions such as the Pacific and mid West contributing in a combined of 5%, maintaining a diversified revenue structures in regional basis. It is worth to mention and that since being in power and serve the smart hardware sector, our revenue also reflects the unique pattern and seasonality of the hardware industry, especially on the international manufacturers represent the results of the business actions and strategies from the past, several quarters or even longer. In Q1, blended gross margin stood about 48.5% with all 3 revenue lines maintained steady margins. Specifically, this quarter's past gross margin rose to 48.4%, primarily due to structural improvements driven by product mix change. Smart solution and SaaS had gross margin of 25.7% and 74.4%, continually to demonstrate the value proposition related through software and technology. In the meantime, our net operation expenses for this quarters were $37.7 million, and nearly 18% decrease from the same period last year, benefiting from a substantial reduction in share-based compensation expenses. A trend that is expected to continue. Excluding equity incentives and other non-operational business factors, our non-GAAP net operation expenses were $29.4 million, down 2% year-over-year. The vast majority of our operation expenses are related to products development and technical teams. Thanks to the groundwork that during our AI transformation since 2023. We have completed -- we have completed the foundational AI product development and service system while maintaining a good discipline in ongoing development. Those structural improvements directly drive the release of operating leverage, enabling us to achieve over $11 million in GAAP net profit in Q1, more than double that of last year’s full year. Non-GAAP net profit, which is $19.3 million are nearly 60% year-over-year increase with a non-GAAP net profit margin of 25.8%. This metrics provides strong support of the company's future business operation, capital expenditures, and shareholder returns. So that includes a brief overview of Q1 financial performance. And also in Q1 is a typical offseason for the industry we serve. So, we really focus our efforts on exploring ideas, opportunities across our product lines. And next, I'll talk about some key moves that we made in our business over the past few months. As we anticipated at the beginning of this year, 2025 marks as a structural transformation year for the entire industry, embrace AI. So, AI is penetrating various industries at decent speed, and for Tuya and the smart sector. This presents a clear window of strategic opportunity for the future. Although, the physical nature of hardware devices means that the commercialization of AI capability on devices is gradually and long-term process. Technical and products reserves in software can take the lead, and then that is precisely our strategy. So, first we are committed to foundationizing and generalizing the AI capability needed in the smarter era to solve the end users' problems and issues. This allows developers to select whatever they needed based on their commercial goals and product designs, thereby promoting the AI adoption. So, focus on 4 core areas. So, large language model integration, hardware development, edge deployment, and open source ecosystems. We release our 4 core engines at the end of April developer conference globally. The Tuya AI agent development platform and the Tuya AI to open and Hedwig. So, the Tuya AI agent development platform significantly reduce the technical threshold for developers to access any mainstream large language model and build various agents functions such as semantic, understandings, and image recognitions. So that is for the agent development. [Tuya AI] [ph], the second one provides a full process to chain from high performance edge AI modeling to [smart card] [ph] deployment, greatly lowering the difficulty and social time for AI hardware development. And to open as our open source framework enables developers to perform differentiated development, innovations and integrate into the ecosystem. The Hedwig platform offers a controllable and secure edge AI computing environment for enterprise customers requiring localized deployment and data platforms and comprises. This entire system forms the technique foundations of our AI platform capabilities that provides a solid support for the largest scale implementation of AI products in the long term. At the same time, we're increasing investment in the development ecosystem -- develop ecosystem around to OS to Open and the T5AI development board. Initially establishing a content plus tools plus community trial frameworks for the developers. Since the developer conference, we have released 8 open-source DIY projects, such as open source desktop test, AI voice boxes, and AI robots. Aimed at building a developer creativity incubation system. Our T5AI developer boards has started to be adapted throughout collaborations with multiple developer platforms and the communities. And our AI developer communities has grown by more than 10,000 new members. Next, we're also committed to identifying the product specific opportunities and driving customers product launches through our smart solution offering directly increasing our revenue and gross profit and achieving high quantity commercial conversions. For example, in the energy domain, we launched the [indiscernible] AI energy assistance with the features strategy -- with the features like the strategy recommendation, power generation forecasting, load authentication, and energy consumption diagnostics, combined with the supporting hardware such as all-in-one energy storage system, circuit breaker, and energy data controllers. It forms a comprehensive integrated hardware software solution, primarily targeting in Europe and Southeast Asia market. In video AI products, we continue to advance the AI capability of smart screens and video devices, focusing on models for detecting and objectives, vehicles, birds and flames combined with the sound recognitions and [animal] [ph] detections to build an integrated hardware and software general production solutions that covers basic home safety needed and extended to semi-commercial scenarios like hotels and real estate. In the meantime, we're also actively promoting external ecosystem collaborations to amplify the enable efforts of the AI platform. For example, in Q1, we partnered with Volcano engines under ByteDance to integrate with Doubao large language model into Tuya AI agent development platform, further enhancing the modeling, understanding and interaction capabilities of our AI agents. In addition, we are collaborating commercially with leading players in retail and marginal -- industries to join development smart hardware such as the AI wearables and home robots, advancing the application of AI in children and family. Collaboration with upstream and downstream partners will further enrich our capabilities and ecosystem scope. Strengthen Tuya's strategies that in a smart sector that creates explanatory projects that global developers and customers can refer and replicate. Although, the AI hardware ecosystem, community, and the product developments are still in the early stage, we believe that explorations in functionality experience and business models will open up space for Tuya and the entire industry's future business exchanges. Laying the foundation of new engines of the long term growth. So, there above is our sharing on Tuya’s Q1 2025 financial performance and the recent company developments. Although there has been partially improvements in external environment recently, uncertainties remains. In response, we will continue to operate diligently and to long term approach of the technique-driven and the platform enabled developments, laying a solid foundation on the AIoT track, continuously capturing industry certainty trends and creating long-term value for Tuya. Finally, a piece of good news. So MSCI recently upgraded -- updated our 2025 ESG rating from single A into a double A. In particularly in the field of secure and compliances to achieve a full score of 10 points, thanks to its extensively compliance experience and through our frameworks. We believe that whether in business, operations, or emerging areas such as the ESG, steadfast progress will bring rewards and recognition. Thank you all operators, and right now we can begin our Q&A section.

Operator:
[Operator Instructions] We will now take our first question from the line of Yang Liu from Morgan Stanley. Please ask your question Yang.

Yang Liu:
[Foreign Language] Let me translate my question. I have 2 questions here. The first one is regarding the AIoT development to have a development platform. Could management share what kind of AIoT hardware is doing well in terms of the shipment or is there any samples that we observed what kind of AI hardware can outperform relatively fast. The second question is regarding the customer's behavior in the past around 2 months, given the huge volatility in tariffs, could you share what the customers doing, in the past 2 months?

Alex Yang:
Thank you Liu. So for the first question, what we see here for the AI sector, we found that there will be 2 major directions that show more interest. The first one is that the audio and the video interaction -- interactive experiences. So one example is bird feeders, we mentioned that in -- I think that in the past quarter. So to be able to help people to recognize the [indiscernible] and to interact with it, and that would be one thing. And the second one -- second typical use cases would be like the toys or the any type of used cases for kids that use the larger language models and capabilities that you can have a new way to interact with the kids, show as education, as entertainment, and as a companion as well. So that would be -- that would be the one. And so recently updates about kit is that in China, we already signed a strategic partnership with [indiscernible], to which we the top tier and the toy group in China, and we're looking forward to help them to launch the production into the market pretty soon -- pretty soon. And so that'd be one thing. So, any type of audio video new interactive experience based on larger language model. And the second is on the data-based dynamic decision making. So one typical used case is will be the energy management system for [indiscernible]. And so through that the AI agents based on the new modeling will be able to continue to not only collect analysis that all the status from either the energy generators, but also from the energy consumption devices, but also in the same time that the agent will be able to come out different types of strategy and turn that into the specific actions and continues to reanalysis the new strategy and improve that. So that can help to do a lot to improve the energy efficiency, no matter it's saving the total consumption, be more greener or saving the energy bill directly. So that will be the 2 very typical trends that we found that to show more interest and have more short-term interactions coming directly from the customers and to the market as well. So that'll be the first question. And the second one, yes, the tariffs become a major topic for the entire market, not only the United States. And so in the past couple of -- in the past 4 or 5 weeks right after the new tariff, topic raises. I think the very direct reaction from the customer is that every people on the international trading pipeline become more conservative and hesitated. Yeah, because nobody knows that what comes out. So right now is that people are waiting for some certainty. And yeah, even now, we have like 90 days of certainty between China and U.S. So people are trying to do some very careful decisions based on the 90 days terms. But I think the longer term I think we need to wait for the final answers between the national negotiations. But I have to mention that for the short term, every people are slowing down or become more conscious about any proactive or aggressive decisions. But to the unique business model and radicalization for Tuya is that the first one, and the tariffs and not [indiscernible] on us because what we deliver will be a portion of key component and technology, especially the software of the finished goods. So, the tariffs are more directly related to my customers, either my brand customers or my manufacturer customers. So that'll be one thing. So, we are like the indirected, and indirectly connected. And the second one is that, and so for us in the past couple of years, what we have been doing is that we just follow the flow that if there's the trend that the entire global economy are we localizing the manufacturing supply chain, we just follow the customers to do that. So not have to stick to one or some specific manufacturers in specific locations. So right now, the manufacturers already cover over countries out there. That'd be the second one. And the third one is that so the tariffs are majorly focused on the finished goods, and the situation is only portion of the cost of that goods. So, that would be -- that's one. But for the tariffs were more, for us that we will pay more attention on the macro impact. So, starting from the buying forces, demand impact on the U.S. market. So, whether that will bring into the retail price raising. And to lower the buying forces for the end user of the American products and also that whether that kind of single nation economy impact is going to be bringing a wider multinational that the macro economy, slowing down or recessions. So, I think that'll be the more long term.

Yang Liu:
Got it, thank you.

Operator:
[Operator Instructions] We will now take the next question from the line of Timothy Chow from Goldman Sachs. Please ask your question, Timothy.

Timothy Chow:
[Foreign Language] Thank you very much for taking my question and congrats on the very strong quarterly results. Two questions here. The first question is on the AI monetization, wondering how, in which area that we would like to plan to monetize the opportunities bring by AI and what is the difference in terms of the pricing model and cost between the new AIoT product services versus the traditional IoT PaaS services that you provide. And secondly is on your full year and second quarter outlook on revenue and profitability. Could management share any color and specifically on the IoT PaaS segment, we have seen very steady upward trend in gross margin of this segment. Could management provide an outlook for this segment as well.

Alex Yang:
Yeah, thank you. I think that for the first question, it's a good one. That's all my customers been asking at the end of last year. So, what we are offering for the -- for the AI capabilities to the customers that we see is the integrate the AI capability into my existing 3 business model. So I still offer that as PaaS SaaS as a solution. So for the customers will be very, very friendly, that they don't, they would -- they don't have to be educated for a new type of business model and how can mix that with their existing one. So it's like a seamless transition. And for the AI income that even I offer as maybe it's AI empowered new PaaS or AI new SaaS that we will consider some of the way pricing. Some example is that recently we offered for the audio interacted AI based on the LLMs, so the PaaS pricing will be different than the regular, my regular PaaS in the same categories. So that would be one. So we can look for that either we use the seamless AI offering to speed up the penetrating of the entire smart devices penetration in the global market, or we use that to re-pricing some of the offerings. So that'd be one thing. And so also that will be very easy to continue to enlarge our partnership with option providers, because right now the major like the Latin language model are providing through the major cloud [indiscernible]. So it's the same way like how we cooperate with them on the cloud infrastructures right now, we cooperate with them in the same model with a larger language model. So that means the first answer for the first question. On the second one, yes, that's combined with the tariffs. Even we don't do the -- we don't involve in the international supply chain or trading business directly, but we're very deeply integrated into that part. We are a core part of that. So the tariffs indeed bring the short time bumping about the demand cycle. So, I can describe that very typically in the major weeks of April, that for all those kind of order shipments to the United States, and a lot of people, most people put a pause button on that and wait and see until that the 2 countries make agreement. So, what we're seeing that this will have topics on globally, majorly from China, U.S., but also slight impact into the Southeast Asia and Mexico, and part of Europe as well, so that type of tariff topics will bring a bumping right. And in the short term, maybe in a quarter or 2. It depends on that how soon that we can make agreement and provide the certainty for the commercial -- for the commerce world, right? All the businessmen are waiting to see that how -- what type of price that they can pay extra and how they can observe that. So for the short term, there will be a bumping, but it's not the first time we experienced that. So this kind of new tariff terms and challenge of business took place since 2018. So, but this seems like the harsher bumping point. So for the short term, there's still uncertainty for that. And even U.S. market only covers less than 10% of overall business, but yeah, that's still 10%. So that's uncertainty and, it's the same for us and my customers. We just wait and see and every people are trying to take more careful actions and decisions, like every single day or week. Yeah, so that's the way we see. And then challenge exists. And but for the long term that we believe that in a year or next year that until, I mean, right after the negotiations being settled, and the internal demand and penetrations will put it back into the normal mode, and all the buying cycles and GTMs will get into the normal cycles again. So that would be what we see for the short term. And for the PaaS margin, I think that the PaaS historical records already proven that we take the high value propositions of the entire industrial cycle. And we are good with it and we are satisfied with that. So for us is that we don't have a strong -- we don't think it will be very necessary. We have to continue to push the margin into a higher cycle, because we're already being the highest -- in the entire one, including my suppliers and my customers. We really take a very high position for that. So, we think that the key part is that to maintain the position, but also in the same time be able to motivate either my suppliers or my customers to continue to invest on these sectors. I think that will be more important for us, because there is an activity will ensure that we continue to improve the penetrations, continues to find more innovations to turn that into reality and bring that into the market. So for the PaaS, we think that it will be fine for us just to maintain at the right range. And you know for us like how we do that in the past 2 or 3 years, so we're satisfied with that. We don't have to ensure being pushed to the next level.

Timothy Chow:
Thank you, that’s very helpful.

Operator:
[Operator Instructions] We will now take our next question from Kai Xiao from CICC. Please ask your question Kai.

Kai Xiao:
[Foreign Language] Thanks very much and congratulations for the strong quarter. I have 2 questions. First, could you share your, the progress of your cooperation with Singapore and Chery and any similar projects can we expect in the coming quarters. And second one is, could you share the geography, expansion of Tuya’s AI developers? And is there any difference on the development of AI applications between domestic markets and overseas markets.

Alex Yang:
Yeah, so yeah, thank you for the questions. So the first one is talking about the The Singapore use cases. I like to put it in the wider coverage. So Singapore use case will be a typical progress we like to do in places for the commercial and or industrial IoT solutions. So the partnership we had or the contract we signed with the with the Singapore government is from the HDB, so it's Singapore Home Development Bureau, and the managing over 1.4 million Singapore public apartments. So majority of Singapore citizens living in the government provided rental apartments, and HDB managed that. And what we're offering is that it's a centralized energy management platform that helps them to able not only censoring the different consumptions in -- in each of the apartments that might contribute significant consumption of the total Singapore nation powers. And also we offer them some controller. So, it helps them to identify how each of the households been using the power to what type of devices, by what purpose, and then through some AI enabled modeling and the control actions to reduce that. And so our over 1.5 years POC with HDB before we sign the contract, that there will be significant reductions about the total consumption of the powers, which can help the Singapore government to reach they want to be the pioneer of ESG nation or smart nation. So, that will be one. And the HDB's plan is that -- we are finalizing the deployment, implementation of the entire platforms, and we will be starting to produce the first shipment of the devices, it's energy hub in the house -- in the home. And so the first implementation about the software platform and the hardware into the first part of the households were taken places around the end of Q3. And their plan is to be able to finish the Phase 1 entire penetration of the homes in 3 years. So that'd be one stuff and we are very excited about it, because that will be showed that, not only providing DIY devices to penetrating homes improvement through those kind of decentralized devices, but we should be able to provide centralized solutions through the right channels like HDBs to have entire regional end users to improve their lives. Yeah, so that'd be one and also we found that use Singapore as a perfect use cases spot, we can -- we will be able to duplicate that and to influence that in other Southeast Asian countries, like the rent apartments, like what we call spatial management solutions, and already show a good potential of growth in Southeast Asia. Yeah, so that's one thing and not only for the apartment management, but also we're really penetrate quite well in some of the telecoms and players in that area too. Yeah, so that'd be one thing. And also in the same time, 2 weeks later, we will open our new Singapore offices and by providing a more local resources to supporting our customers, penetrations, implementations, but also in the same time to search for all business opportunities in that area. And expanding into global market, yeah, so that would be the answers for the Singapore parts. Could you repeat the question? So the second one is about returns right, the shareholder returns, right?

Kai Xiao:
The expansion of AI developers.

Alex Yang:
So AI developers what we hear is that I think it's just started I think right after the DeepSeek expansion in January. And so the first one is that we found majority of my customers and developers are reaching out to back to us to asking about all the AI capability offering, because in the past, on the recognition side is that they know AI is there, and a lot of news about AI, but they don't really think that the AI offering will be connected to their existing business. Either it's more expensive, too expensive, unaffordable, or it would be too difficult. So, I think that right after the DeepSeek exposure, both in America and the China market, every customer aware that they can bring AI into the physical use cases. And then I think the first one is that we got so many calls. And then we use that time windows to launch either launch a new platform, the AI agent development platform, but also to hold our -- to launch that as our developer conference. So that brings a very positive and very active feedback. And so right now the AI developer penetration is that we're trying to educate those customers who call and by telling them what capabilities out there in the platform and how they can develop upon those capabilities to turn that into the real devices use cases. It's more like an educating period. And right now I will see that the demand comes from the global market. So majority of my major or key customers either from Europe, from America, from Latin America and Southeast Asia. So they are -- those customers have very strong retail channels, or have very strong brands, who can do quite well about the end users' educations and marketing. So those customers will be the major cause and on the demanding side. And on the technical development side, so in 2 parts. The first part is that those ones who are focused on turn the AI or in-build the AI capability into the device. So it's more following the right penetration on the device development and manufacturing. So maybe a bigger portion come from mainland China, but still we got multiple ones from Korea, from Turkey, from Southeast Asia too. And so that's following -- perfectly fits the penetration about the hardware manufacturing deployment. That's the first one. And the second one is about software. So software, I'll say that it's more diversified. And either it's different types of software innovators through the MCP protocols, and more come from the mobile internet developer ecosystem and also the cloud-based developers and who come through the cloud infrastructures into us. Yeah, so it's more diversified on a global basis. I think that's for the first quarter or including till now, it's still in the Phase 1. It's about the recognition about capability, education about the developer kit, and then a lot of prototype, POC and demonstration deployment.

Operator:
We will now take our next question from the line of John Roy from Water Tower Research. Please ask your question, John.

John Roy:
Yeah, I have a slightly different question. I was wondering if you could give us any insight into how you're using AI internally, internal processes, documents, code development, etc. And if that can really possibly increase margins and you know decrease costs going forward?

Alex Yang:
Yeah, thank you, John. So, I think that every single department within the company is trying to using different types of AI tools. I can take some of these examples, right? Like for marketing team right now that the majority of all the marketing resources we use are generated through AI. So we don't require a lot of labor-based designer to do that. And including our operation team to do all those kind of online training webinars. I think a big portion of the webinar contents are generated through AI too. And also we have our TikTok live view to the end users, the AI helps a lot too, and that's the marketing side and including the legal and human resources side we're using AI to help us either to do the research and review on the -- the candidate's resume and with keywords that we identify and also to help us to double check with all those standardized contractors come directly from the customer side to see that whether there's any modifications, etc. So use that to improve the efficiency about the internal processing of either the HR and the legal process. So that'd be one thing and another use case is about the development for sure. That we're using AI toolkit, different [indiscernible] kit to do the project management and analyzing about the progress of the development and the coding, double checking, debugging, pre-setting, and maintenance of the cloud infrastructure. So, what we call auto patrolling on the maintenance of the cloud services as well. Yeah, so that would be typical things. We believe that in the future, like we mentioned, not only providing the AI capability to help the customers to build more commercialized or more higher [RI] [ph] products, but also in the same time that we continue to using AI internally to improve our operation efficiency and so either lowering the cost of the operation or to increase the income for our customers to enhance our lab operating leverage. That'd be some basic.

John Roy:
Thank you so much and congratulations on the quarter.

Alex Yang:
Thank you for that. We’re excited about the long term future, because this is the only very starting point for the AI penetration for the physical world, very, very early. We're very excited about that.

Operator:
There are no additional questions at this time. I'll now hand back to the management team for closing remarks.

Alex Yang:
Okay, thank you everyone. Thank you operator, and thanks for all of you for participating on today’s call. See you next quarter and have a nice day today. Bye.

Operator:
Thank you for your participation in today's conference. This does conclude the program. You may now disconnect your lines.