Agora, Inc.

Good day and thank you for standing by. Welcome to the Agora, Inc. First Quarter 2023 Financial Results Conference Call. At this time, all participants are in the listen-only mode. After the speaker's presentation, there will be question and answer session. To ask a question during the session, you will need to press star 1 1 on your telephone. You will then hear an automated message advising your hand is raised. To withdraw your question, please press star 11 again. Please be advised that today's conference is being recorded. And I'd like to hand the conference over to the IR Director, Ms. Fiona Chen. Please go ahead, ma'am.

Thank you, operator. Good morning and evening, everyone. Thank you for joining us for Agarwal Inc.' 's first quarter 2023 earnings conference call. Our earnings results press release earnings presentation and an explanatory note on certain reporting and disclosure adjustments, SEC filings, and a replay of today's call can be found on our IO website at investor.agora.io. Joining me today are Tony Zhao, founder, chairman, and the CEO, Jim Bowen, our CFO. Reconciliations between our GAAP and non-GAAP results can be found in our earnings press release. During this call, we will make forward-looking statements about our future financial performance and other future events and the trends. These statements are only predictions that are based on what we believe today, and the actual results may differ materially. These forward-looking statements are subject to risks, uncertainty, assumptions, and other factors that could affect our financial results and the performance of our business, and which we discuss in detail in our filings with the SEC, including today's earnings press release and the risk factors and other information contained in the final prospectus relating to our initial public offering. Agora Inc. remains no obligation to update any forward-looking statements we may make on today's call. With that, let me turn it over to Tony. Hi, Tony.

Thanks, Fiona. And welcome, everyone, to our earnings call. Before diving into our operational results for the quarter, I would first like to make a few remarks regarding recent changes in the organizational structure of our company. Since our inception nine years ago, we have mostly operated under the Agora brand globally. To better meet the requirements of our customers and compete more effectively in the unique markets we serve, we will now operate two independent divisions under the same holding company. Our Agora division will focus on our business in the US and international markets, and our Shenwang division will focus on our business in the China market. Agora and Shenwang will each be run by a local management team and adopt local standards and best practices. We have appointed Stanley Wei as Chief Operating Officer of Agora and Robin Liu as Chief Operating Officer of Shonwa. We believe this strategic reorganization will allow us to optimally focus our resources on the specific needs and priorities of each business. Agora will focus on acceleration of growth and on gaining market share in the US and international markets, leveraging our leading technology and comprehensive product offerings. Shenwang will focus on enhancing our quality of experience advantage and improving the ease of adoption of our products, further strengthening our competitive position in the China market. By empowering the local management team of each division, we will be able to respond more quickly to the needs of our customers and become more agile as new opportunities emerge. From this quarter onwards, we will report our revenue and operating metrics separately for Agora and Shenwang to help investors better understand the dynamics and our operational results in each of these markets. Our revenue for the first quarter was $15.1 million for Agora, up 10% year-over-year, and $21.3 million for Shenwang, down 14% year-over-year. Agora's revenue growth was primarily due to business extension and usage growth in US and international markets. Shenwang's decrease in revenue was primarily due to the appreciation of US dollar over disposal of E-Small's customer engagement cloud business, and decrease in usage from K-12 academic tutoring customers. After adjusting for these factors, Shengwang's revenue would have increased by 3% compared to the same quarter last year. As of the end of this quarter, We have nearly 1,500 active customers for Agora and nearly 4,000 for Shonla, an increase of 31% and 2%, respectively, compared to one year ago. Now, moving on to our product and technology updates. First, starting with Agora, the flexible classroom of our local ACOS solution designed for building scalable and customizable online classroom has continued to gain momentum in the US and the global market, as the demand for online learning continues to grow. Recently, the flexible classroom was named a finalist by 2023, the EdTech Award in the e-learning blended solution or remote solution category. Previously, we mentioned our partnership with HTC to accelerate the adoption of real-time engagement in VR applications. In HTC's wide ecosystem, developers use our video SDK to power real-time collaboration between end users within virtual environments. Recently, we were recognized as a Webby Award Honorary under the category of Metaverse, immersive and virtual best real-time experience for how our technology is used in HTC's wide SyncVR ecosystem. To further increase the value of our premium and enterprise support package for our customers, we have added integrations with Okta and Datadog. The Okta integration enables customers to manage the team members and access to Aurora console through Okta identity management tools. The DeepDock integration with Agora Analytics enables customers to push analytics data covering usage, quality, and performance directly to the DeepDock platform for analysis and visualization. Moving on to Shonwa, we recently upgraded our real-time karaoke product When this solution was first launched in 2021, it could accommodate two users singing together at the same time. If a third user wanted to join, one of the two current users would have to give up the microphone. This limitation was mainly due to the technology challenge at the time of mixing multiple real-time soundtracks in a highly synchronous manner and delivers output to all users with low latency. With our latest version, up to eight users can now sing together, which unlocks a wide range of new features for our customers' applications. For example, all users in the virtual karaoke room can now participate in a singing battle or take random turns singing part of songs together. empowered by our upgraded product our customers can now replicate in-person priority experience within their application next i would like to discuss the combination of generative ai and real-time engagement and enormous opportunity this combined this combination holds for the future since the end of last year the world has been captivated by the latest advance in large language model and generative AI. People widely believe that generative AI, despite being in its early stage, will drive significant paradigm shifts in many industries. And real-time engagement is no exception. Generative AI can empower to dramatically enhance end-user experience in their applications or create new use cases that were previously impossible. Let's look at some examples. The advent of powerful large language model has significantly improved the performance of chatbots in text format, as can be seen in the growing popularity of chat GPT and AI-powered search engines. However, this is only the beginning. As AI models continue to evolve and gain more powerful multimodal capabilities, they will become more adept at processing voice and video feeds and generating response in real time. Using IoT hardware as input-output devices, people will have an all-powerful human-like AI body that can provide information. carry on tasks, or engage in casual chats all through voice and video. Education is another industry where people have strong hopes that generating AI can help revolutionize experiences. By analyzing student data, AI-powered tutors can generate personalized learning plans and deliver learning content in an interactive manner with everything tailored to best match a student's learning progress, pattern, and real-time feedback. In addition, AI tutoring can also provide affordable and tailored education opportunity to those who previously did not have access to a quality education. There is also huge potential for generative AI to disrupt the social entertainment and gaming industry. A virtual matchmaker can do the job of making initial introduction and facilitating conversations just as well as a real person in a dating application. If you need to leave an interactive session for a short while or being disrupted by network connection, your digital twin can take over and continue the conversation. When you come back, a summary can be immediately available for you to catch up. For online social gaming, it will become almost impossible to tell if your teammates or opponents are real players or AI-controlled non-player characters. Looking at the examples above and many others, we see something in common. Generative AI can significantly expand the scope and opportunity of real-time engagement. Previously, real-time engagement largely took place among groups of people. With the help of large language model, RTE can now occur among people, digital trends of people, and fully virtual AI-based characters. opening the door to a much broader range of possible use cases. In addition, with greatly enhanced user experience, RTE will become more intelligent, immersive, and enjoyable. This trend will likely increase the overall usage of real-time engagement solutions multiple times and bring more business opportunities. We have been working closely with our customers to create pilot applications in certain verticals, and we will continue to monitor the latest developments while assisting our customers in utilizing generative AI in their real-time engagement use cases. Before concluding my prepared remarks, I want to thank both Agora and Shenglong team for their hard work during this transitional period. I believe this strategic reorganization will sharpen our identity and strengthen our position in both the global and China markets. And most importantly, allow us to serve developers and customers in a more agile and efficient way. With that, let me turn things over to Jingbo, who will reveal our financial results.

Thank you, Tony. Hello, everyone. Let me start by first discussing certain reporting and disclosure adjustments in our financial results. Then I will review financial results for the first quarter and the outlook for the second quarter of 2023. Following our recent reorganization, Agora Inc. is now the holding company of two independent businesses, Agora and Shunwo. which will operate under their own unique brands and distinct legal entities and will be run by separate local management teams. Beginning from the first quarter of 2023, we will report revenues separately for Agora and Xiongwa based on the legal entities with which customers enter into contracts. This differs from our previous practice, which was to separately report revenues for China and US and international based on geographies of usage. For example, certain Chinese customers offer the application that primarily target end users outside China. Such revenues were previously included under US and international. We now include such revenues under to reflect the of contractual relationships. For the same reason, we will report number of active customers and dollar-based net retention rate, or DBNRR, separately for Agora and XiongWang. But definition of active customers remain unchanged. The calculation methodology of retention rate is same as expansion rate, which is the term we previously used. As almost all revenues generated from Agora customers are denominated in U.S. dollar and almost all revenues generated from ShenWang customers are denominated in RMB. We calculate DB and RR in U.S. dollar for Agora and in RMB for ShenWang instead of converting everything into U.S. dollars. Revenues for Eastmob's chat API business will also be included in calculating active customers. and retention rate for show-off business. A detailed explanatory note, including a recast historical results reflecting these adjustments, can be found on the Investor Relations website. Moving on to our financial results for Q1. Agora revenues were $15.1 million in the first quarter of 2023, an increase of 10.2% year-over-year and a decrease of 4.4% quarter-over-quarter. The year-over-year increase was primarily due to our business expansion and usage growth. The quarter-over-quarter decrease was primarily due to challenging macroeconomic environment. As we mentioned in previous earnings costs, the interest rate hikes Worldwide inflationary pressure and tightening of venture capital funding starting from the second half of last year had negatively impacted some of our customers' business and financial conditions and their ability to raise funding, which led to reduced usage of our products and increased pricing sensitivity. Show on revenues were $21.3 million in the first quarter of 2023, a decrease of 14.5% year-over-year, and a decrease of 12.3% quarter-over-quarter. The year-over-year decrease was primarily due to the disposal of eSmart's Customer Engagement Cloud business, a decrease in usage from K-12 academic tutoring sector following regulatory change, and the depreciation of RMB against US dollar. If we exclude these factors, revenue denominated in RMB would have increased 2.7% year over year. The quarter over quarter decrease was primarily due to the disposal of customer engagement cloud business, lower usage during Chinese New Year holiday for internet customers, and longer than expected sales cycle for traditional enterprise customers. Dollar-based net retention rate for Gora is 130%. Dollar-based net retention rate for Shenwang is 92%, excluding revenues from the K-12 academic tutoring sector. Moving on to cost and expenses. For my following comments, I will focus on non-GAAP results, which exclude share-based compensation expenses, acquisition-related expenses, financing related expenses, amortization expenses of acquired intangible assets, income tax related to acquired intangible assets, and impairment of goodwill. Non-GAAP gross margin for the first quarter was 63.3%, which was 0.3% higher than Q1 2022, mainly due to the disposal of customer engagement cloud business, which had lower gross margin. As we mentioned in our previous earnings calls, we restructured and reduced our global workforce in Q4 2022. As a result, non-GAAP R&D expenses were 17.4 million in Q1, a decrease of 31.2% year over year. Non-GAAP R&D expenses represented 47.8% of total revenues in the quarter compared to 65.6% in Q1 last year. Non-GAAP sales and marketing expenses were 8.5 million in Q1, a decrease of 26.3% year-over-year. Sales and marketing expenses represented 23.4% of total revenues in the quarter compared to 30% in Q1 last year. Non-GAAP G&A expenses were 6.9 million in Q1, a decrease of 7% year-over-year. G&A expenses represented 18.8% of total revenues in the quarter compared to 19.1% in Q1 last year. Non-GAAP operating loss was 9.2 million, translating to a 25.4% non-GAAP operating loss margin for the quarter, compared to an operating loss margin of 49% in Q1 last year. Adjusted EBITDA was an active 6.4 million, translating to a 17.7% adjusted EBITDA loss margin for the quarter, significantly lower than the loss margin of 42.6% in Q1 last year. Investment loss was 4.4 million in Q1, primarily due to the fair value change in equity investments of 2.9 million, credit loss in debt investments of 1.2 million, as well as disposal loss of 0.3 million. Now turning to cash flow. Operating cash flow was negative 8.9 million in Q1 compared to negative 15.9 million last year. Free cash flow was negative 9.1 million compared to negative 17 million last year. Moving on to balance sheet. We ended Q1 with 416.5 million in cash, cash equivalents. bank deposits, and financial products issued by banks. Net cash flow, net cash outflow in the quarter was mainly due to free cash flow of an active 9.1 million, share repurchase of 19.4 million, and cash paid in relation to headquarters project of 5.1 million. During Q1, we repurchased approximately 21.6 million of our Class A ordinary shares, equivalent to 5.4 million ADS for $19.5 million, representing 10% of our 200 million share repurchase program. As of the end of Q1, we had in aggregate repurchased approximately 57.4 million of our Class A ordinary shares, equivalent to 14.4 million ADS for $61.3 million, representing 31% of our share repurchase program. Now turning to guidance. Recently, we saw increased uncertainties in macroeconomic conditions. such as foreign exchange rate, inflation, and interest rates, which may have greater than expected impact on our customers' business, and in turn, add more uncertainty to our revenues. Therefore, starting this quarter, we'll be providing quarterly revenue guidance instead of four-year guidance. For the second quarter of 2023, we currently expect total revenues to be in the range of $34 and $37 million. This forecast reflects our current and preliminary views on the market and operational conditions, which are subject to change. I want to express my deepest appreciation to both Agora and the Shun Wang teams for your hard work during this challenging period, and to our investors for your trust in our vision and our team. Thank you, everyone, for attending the call today. Let's open it up for questions.

Thank you. As a reminder to ask a question, please press star 11 on your telephone and wait for your name to be announced. To withdraw your question, please press star 11 again. Please stand by while we compile the Q&A roster. Once again, that's star 11 for questions. Our first question comes from the line of Yang Liu from Morgan Stanley. Please go ahead, Yang.

Thanks for the opportunity to ask questions. Two questions from my side. The first one is about the outlook. The number implies a larger fetish of sequential decline in second quarter. Could management share more about the outlook for Shengwang and Agora differently for the second quarter? Do you see further diverging of the growth trend or similar trend in the next few quarters for the two entities? And my second question is regarding the AI impact to the company. Tony mentioned a lot in terms of the product and the customer demand or use case driven by the new AIGC technology. But I'm wondering if there's any potential cost-saving room from the new technology. Thank you.

All right, I'll discuss the demand from both markets. On Shenglong's side, we can broadly push Shenglong's customers into three segments. One is domestic digital native and another is domestic digital transformation and going overseas customers. On digital native customers, Recently, demand remained soft, primarily due to macroeconomic challenges and regulation. However, on this segment, as I mentioned earlier, we also see large opportunities on generative AI-powered use cases. That would take some time. On digital transformation side, recently we saw longer sales cycle, but there is clearly a lot of unmet demand from traditional enterprises. We expect this segment to generate sustained growth for us. On going overseas customers, the overall growth momentum has been strong. On Agora side, we see overall positive demand momentum in more developed market, especially on the following verticals. Live streaming e-commerce, creator and finance economy, sports live streaming, future of work. For emerging market, we see pressure on our customers due to macroeconomic challenges and market competition, such as in South Asia and Southeast Asia. That's from the outlook of the demand side. And generative AI side, We actually have invested in this area for quite some time to help improve the audio video quality. As I mentioned, a large language model can actually help create more use cases in many verticals. If your question is around reduced cost, I don't think it's going to hugely help us to reduce cost. But in terms of disrupting many business models, like in education, AI-powered tutor can be much cheaper than a real teacher. Or virtual celebrity or singer can also be much cheaper, or virtual matchmaker, the same sense, or even in social space. As you can see, a popular character who's busy talking to people in one social platform can be also an AI-powered person to reduce the platform's cost to maintain and pay a real person who's active on that platform. I think on those, it can hugely disrupt their business.

Thank you.

Right, thank you. Our next question comes from the line of Daily Lee from Bank of America Securities. Please ask your question, Daily.

Hi, thanks for taking my question. I have one question regarding our international, the Agora business. For this quarter, it seems delivered quite solid growth. I know what's the take drivers for the growth, for example, regarding the volume and ASP, and what's the outlook for the volume and ASP going forward? Thank you.

As we explained in previous earnings calls, so the market in China and the market in U.S. and other international markets at different stage of development in terms of adopting the RT technology. So China market is comparatively more mature. So in the past 18-24 months, we certainly see US market catching up quite rapidly. For example, the live streaming e-commerce use case. that was already popular in China two years ago, and probably even earlier. But that has really just taken off in the past 12, 18 months in the US. And that has driven a lot of new demand for technology. Also, we have been expanding into more geographic regions, South Asia, Middle East, and even South America. So that explains the stronger growth momentum on the agrarian side. And looking forward, as Tony just explained, in China we do see more macro challenges, the overall macroeconomic environment, and also regulation. In terms of the more traditional enterprise customers in China, we also see more budget constraints. So looking at the next few quarters, we are also more optimistic in terms of demand for the Acura side.

Thank you, Jinbozong.

Thank you. Our next question comes from the line of Ethan Zhang from Nomura. Please ask your question, Ethan.

Hi, thank you management for taking my question. This is Bing Duan from Nomura. I have one question regarding the China market as we noted that There are a few changes in the competition landscape, such as AliCloud's plan of spinoff and the recent price cut from the large cloud companies like Alibaba and also some China's telecom operators. So how do we see this impact our business in China, such as the ASP and margins, and also about our future market to share in China? Thank you.

First of all, the public cloud price adjustment recently does not really you know, relate to us and giving pressures to us. However, in China, as I mentioned earlier, the overall competition has remained strong for a long time and the demand is actually, especially for domestic native part, is actually soft. You know, with this situation, you know, we do practically enhance our competitive edge Sometimes we reduce our pricing on certain areas to increase competitive pressure and try to gain more market share. For major public cloud, previously also rolling out similar service in RTE sector. Honestly, there's no one being left. Everyone already released some similar product already. course of past two, three years, many of them have mostly exited the business. They either stopped really selling that product or some already started a partnership with us and available our product in their offerings.

Thank you, Tony.

Thank you.

As a reminder, to ask a question, please press star 1 1 on your telephone.

To ask a question, please press star 1 1.

I am showing no further questions. I'll now turn the conference back to Fiona for closing remarks.

Thank you, operator.

Thank you, everybody, for joining our call today. Again, if there are any further questions, feel free to contact us. And also, the presentation and the remarks of this call will be posted on our website. Thank you, everybody. Thank you.

Thank you. This concludes today's conference call. Thank you for participating. You may now disconnect.

Thank you for that.

Thank you. Thank you. Thank you. Thank you.

Good day and thank you for standing by. Welcome to the Agora Inc. First Quarter 2023 Financial Results Conference call. At this time, all participants are in the listen-only mode. After the speaker's presentation, there will be question and answer session. To ask a question during the session, you will need to press star 1 1 on your telephone. You will then hear an automated message advising your hand is raised. To withdraw your question, please press star 11 again. Please be advised that today's conference is being recorded. And I'd like to hand the conference over to the IR Director, Ms. Fiona Chen. Please go ahead, ma'am.

Thank you, operator. Good morning and evening, everyone. Thank you for joining us for Agarwal Inc.' 's first quarter 2023 earnings conference call. Our earnings results press release earnings presentation and an explanatory note on certain reporting and disclosure adjustments, SEC filings, and a replay of today's call can be found on our IO website at investor.agora.io. Joining me today are Tony Zhao, founder, chairman, and the CEO, Jim Bowen, our CFO. Reconciliations between our GAAP and non-GAAP results can be found in our earnings press release. During this call, we will make forward-looking statements about our future financial performance and other future events and the trends. These statements are only predictions that are based on what we believe today, and the actual results may differ materially. These forward-looking statements are subject to risks, uncertainty, assumptions, and other factors that could affect our financial results and the performance of our business, and which we discuss in detail in our filings with SEC, including today's earnings press release and the risk factors and other information contained in the final prospectus relating to our initial public offering. Agora Inc. remains no obligation to update any forward-looking statements we may make on today's call. With that, let me turn it over to Tony. Hi, Tony.

Thanks, Fiona. And welcome, everyone, to our earnings call. Before diving into our operational results for the quarter, I would first like to make a few remarks regarding recent changes in the organizational structure of our company. Since our inception nine years ago, we have mostly operated under the Agora brand globally. To better meet the requirements of our customers and compete more effectively in the unique markets we serve, we will now operate two independent divisions under the same holding company. Our Agora division will focus on our business in the US and international markets, and our Shenwang division will focus on our business in the China market. Agora and Shenwang will each be run by a local management team and adopt local standards and best practices. We have appointed Stanley Wei as Chief Operating Officer of Agora and Robin Liu as Chief Operating Officer of Shonwa. We believe this strategic reorganization will allow us to optimally focus our resources on the specific needs and priorities of each business. Agora will focus on the acceleration of growth and on gaining market share in the US and international markets, leveraging our leading technology and comprehensive product offerings. Shenwang will focus on enhancing our quality of experience advantage and improving the ease of adoption of our products, further strengthening our competitive position in the China market. By empowering the local management team of each division, we will be able to respond more quickly to the needs of our customers and become more agile as new opportunities emerge. From this quarter onwards, we will report our revenue and operating metrics separately for Agora and Shenwang to help investors better understand the dynamics and our operational results in each of these markets. Our revenue for the first quarter was $15.1 million for Agora, up 10% year-over-year, and $21.3 million for Shenwang, down 14% year-over-year. Agora's revenue growth was primarily due to business extension and usage growth in U.S. and international markets. Shenwang's decrease in revenue was primarily due to the appreciation of U.S. dollar over disposal of ESMOS customer engagement cloud business, and decrease in usage from K-12 academic tutoring customers. After adjusting for these factors, Shenwang's revenue would have increased by 3% compared to the same quarter last year. As of the end of this quarter, We have nearly 1,500 active customers for Agora and nearly 4,000 for Shonwa, an increase of 31% and 2%, respectively, compared to one year ago. Now, moving on to our product and technology updates. First, starting with Agora, the flexible classroom of our local ACOS solution designed for building scalable and customizable online classroom has continued to gain momentum in the US and the global market, as the demand for online learning continues to grow. Recently, the flexible classroom was named a finalist by 2023, the EdTech Award in the e-learning blended solution or remote solution category. Previously, we mentioned our partnership with HTC to accelerate the adoption of real-time engagement in VR applications. In HTC's wide ecosystem, developers use our video SDK to power real-time collaboration between end users within virtual environments. Recently, we were recognized as a Webby Award owner under the category of Metaverse, immersive and virtual, best real-time experience for how our technology is used in HTC's wide SyncVR ecosystem. To further increase the value of our premium and enterprise support package for our customers, we have added integrations with Okta and Datadog. The Okta integration enables customers to manage the team members and access to Aurora console through Okta identity management tools. The DeepDock integration with Agora Analytics enables customers to push analytics data covering usage, quality, and performance directly to the DeepDock platform for analysis and visualization. Moving on to Shonwa, we recently upgraded our real-time karaoke product When this solution was first launched in 2021, it could accommodate two users singing together at the same time. If a third user wanted to join, one of the two current users would have to give up the microphone. This limitation was mainly due to the technology challenge at the time of mixing multiple real-time soundtracks in a highly synchronous manner and deliver the output to all users with low latency. With our latest version, up to eight users can now sing together, which unlocks a wide range of new features for our customers' applications. For example, all users in the virtual karaoke room can now participate in a singing battle or take random turns singing part of songs together. empowered by our upgraded product, our customers can now replicate in-person quality experience within their application. Next, I would like to discuss the combination of generative AI and real-time engagement and the enormous opportunity this combination holds for the future. Since the end of last year, the world has been captivated by the latest advance in large language model and generative AI. People widely believe that generative AI, despite being in its early stage, will drive significant paradigm shifts in many industries. And real-time engagement is no exception. Generative AI can empower to dramatically enhance end-user experience in their applications or create new use cases that were previously impossible. Let's look at some examples. The advent of powerful large language model has significantly improved the performance of chatbots in text format, as can be seen in the growing popularity of chat GPT and AI-powered search engines. However, this is only the beginning. As AI models continue to evolve and gain more powerful multimodal capabilities, they will become more adept at processing voice and video feeds and generating response in real time. Using IoT hardware as input-output devices, people will have an all-powerful human-like AI body that can provide information. carry on tasks, or engage in casual chats all through voice and video. Education is another industry where people have strong hopes that generating AI can help revolutionize experiences. By analyzing student data, AI-powered tutors can generate personalized learning plans and deliver learning content in an interactive manner with everything tailored to best match a student's learning progress, pattern, and real-time feedback. In addition, AI tutoring can also provide affordable and tailored education opportunity to those who previously did not have access to a quality education. There is also huge potential for generative AI to disrupt the social entertainment and gaming industry. A virtual matchmaker can do the job of making initial introduction and facilitating conversations just as well as a real person in a dating application. If you need to leave an interactive session for a short while or being disrupted by network connection, your digital twin can take over and continue the conversation. When you come back, a summary can be immediately available for you to catch up. For online social gaming, it will become almost impossible to tell if your teammates or opponents are real players or AI-controlled non-player characters. Looking at the examples above and many others, we see something in common. Generative AI can significantly expand the scope and opportunity of real-time engagement. Previously, real-time engagement largely took place among groups of people. With the help of large language model, RTE can now occur among people, digital trends of people, and fully virtual AI-based characters. opening the door to a much broader range of possible use cases. In addition, with greatly enhanced user experience, RTE will become more intelligent, immersive, and enjoyable. This trend will likely increase the overall usage of real-time engagement solutions multiple times and bring more business opportunities. We have been working closely with our customers to create pilot applications in certain verticals, and we will continue to monitor the latest developments while assisting our customers in utilizing generative AI in their real-time engagement use cases. Before concluding my prepared remarks, I want to thank both Agora and Shenglong team for their hard work during this transitional period. I believe this strategic reorganization will sharpen our identity and strengthen our position in both the global and China markets. And most importantly, allow us to serve developers and customers in a more agile and efficient way. With that, let me turn things over to Jingbo, who will reveal our financial results.

Thank you, Tony. Hello, everyone. Let me start by first discussing certain reporting and disclosure adjustments in our financial results. Then I will review financial results for the first quarter and the outlook for the second quarter of 2023. Following our recent reorganization, Agora Inc. is now the holding company of two independent businesses, Agora and Shunwo. which will operate under their own unique brands and distinct legal entities and will be run by separate local management teams. Beginning from the first quarter of 2023, we will report revenues separately for Agora and Shenhua based on the legal entities with which customers enter into contracts. This differs from our previous practice, which was to separately report revenues for China and US and international based on geographies of usage. For example, certain Chinese customers offer the application that primarily target end users outside China. Such revenues were previously included under US and international. We now include such revenues under to reflect the of contractual relationships. For the same reason, we will report number of active customers and dollar-based net retention rate, or DBNRR, separately for Agora and XiongWang. But definition of active customers remain unchanged. The calculation methodology of retention rate is same as expansion rate, which is the term we previously used. As almost all revenues generated from Agora customers are denominated in US dollar, and almost all revenues generated from ShenWang customers are denominated in RMB. We calculate DB and RR in US dollar for Agora and in RMB for ShenWang, instead of converting everything into US dollars. Revenues for Eastmob's chat API business will also be included in calculating active customers and retention rate for show-on business. A detailed explanatory note, including a recast historical results reflecting these adjustments, can be found on Investor Relations website. Moving on to our financial results for Q1, agrarian revenues were $15.1 million in the first quarter of 2023, an increase of 10.2% year-over-year, and a decrease of 4.4% quarter-over-quarter. The year-over-year increase was primarily due to our business expansion and usage growth. The quarter-over-quarter decrease was primarily due to challenging macroeconomic environment. As we mentioned in previous earnings calls, the interest rate hikes, worldwide inflationary pressure, and tightening of inter-capital funding starting from the second half of last year had negatively impacted some of our customers' business and financial conditions and their ability to raise funding, which led to reduced usage of our products and increased pricing sensitivity. Show on revenues were $21.3 million in the first quarter of 2023, a decrease of 14.5% year-over-year, and a decrease of 12.3% quarter-over-quarter. The year-over-year decrease was primarily due to the disposal of E-SMAP's customer engagement cloud business, a decrease in usage from K-12 academic tutoring sector following regulatory change, and the depreciation of RMB against US dollar. If we exclude these factors, revenue denominated in RMB would have increased 2.7% year over year. The quarter over quarter decrease was primarily due to the disposal of customer engagement cloud business, lower usage during Chinese New Year holiday for internet customers, and longer than expected sales cycle for traditional enterprise customers. Dollar-based net retention rate for Gora is 130%. Dollar-based net retention rate for Shenwang is 92%, excluding revenues from the K-12 academic tutoring sector. Moving on to cost and expenses. For my following comments, I will focus on non-GAAP results, which exclude share-based compensation expenses, acquisition-related expenses, financing related expenses, amortization expenses of acquired intangible assets, income tax related to acquired intangible assets, and impairment of goodwill. Non-GAAP gross margin for the first quarter was 63.3%, which was 0.3% higher than Q1 2022, mainly due to the disposal of customer engagement cloud business, which had lower gross margin. As we mentioned in our previous earnings calls, we restructured and reduced our global workforce in Q4 2022. As a result, non-GAAP R&D expenses were 17.4 million in Q1, a decrease of 31.2% year over year. Non-GAAP R&D expenses represented 47.8% of total revenues in the quarter compared to 65.6% in Q1 last year. Non-GAAP sales and marketing expenses were $8.5 million in Q1, a decrease of 26.3% year-over-year. Sales and marketing expenses represented 23.4% of total revenues in the quarter compared to 30% in Q1 last year. Non-GAAP G&A expenses were 6.9 million in Q1, a decrease of 7% year-over-year. G&A expenses represented 18.8% of total revenues in the quarter compared to 19.1% in Q1 last year. Non-GAAP operating loss was 9.2 million, translating to a 25.4% non-GAAP operating loss margin for the quarter, compared to an operating loss margin of 49% in Q1 last year. Adjusted EBITDA was an active 6.4 million, translating to a 17.7% adjusted EBITDA loss margin for the quarter, significantly lower than the loss margin of 42.6% in Q1 last year. Investment loss was 4.4 million in Q1, primarily due to the fair value change in equity investments of 2.9 million, credit loss in debt investments of 1.2 million, as well as disposal loss of 0.3 million. Now turning to cash flow. Operating cash flow was negative 8.9 million in Q1 compared to negative 15.9 million last year. Free cash flow was negative 9.1 million compared to negative 17 million last year. Moving on to balance sheet. We ended Q1 with 416.5 million in cash, cash equivalents. bank deposits, and financial products issued by banks. Net cash flow, net cash outflow in the quarter was mainly due to free cash flow of an active 9.1 million, share repurchase of 19.4 million, and cash paid in relation to headquarters project of 5.1 million. During Q1, we repurchased approximately 21.6 million of our Class A ordinary shares, equivalent to 5.4 million ADS for $19.5 million, representing 10% of our 200 million share repurchase program. As of the end of Q1, we had in aggregate repurchased approximately 57.4 million of our Class A ordinary shares, equivalent to 14.4 million ADS for $61.3 million, representing 31% of our share repurchase program. Now turning to guidance. Recently, we saw increased uncertainties in macroeconomic conditions, such as foreign exchange rate, inflation, and interest rates, which may have greater than expected impact on our customers' business, and in turn, add more uncertainty to our revenues. Therefore, starting this quarter, we'll be providing quarterly revenue guidance instead of full-year guidance. For the second quarter of 2023, we currently expect total revenues to be in the range of $34 and $37 million. This forecast reflects our current and preliminary views on the market and operational conditions, which are subject to change. In closing, I want to express my deepest appreciation to both Agora and Shunwang teams for your hard work in this challenging period, and to our investors for your trust in our vision and our team. Thank you everyone for attending the call today. Let's open it up for questions.

Thank you. As a reminder, to ask a question, please press star 11 on your telephone and wait for your name to be announced. To withdraw your question, please press star 11 again. Please stand by while we compile the Q&A roster. Once again, that's star 11 for questions. Our first question comes from the line of Yang Liu from Morgan Stanley. Please go ahead, Yang.

Thanks for the opportunity to ask questions. Two questions from my side. The first one is about the outlook. The number implies a largely sluggish or slightly sequential decline in second quarter. Could management share more about the outlook for Shengwang and Agora differently for the second quarter. Do you see further diverging of the growth trend or similar trend in the next few quarters for the two entities? And my second question is regarding the AI impact to the company. Tony mentioned a lot in terms of the product and the customer demand. or use case driven by the new AIGC technology. But I'm wondering if there's any potential cost-saving room from the new technology. Thank you.

All right. I'll talk about, discuss the, you know, the demand from both market. On Shen Wang's side, we can broadly push on the customer into three segments. One is domestic digital native, and another is domestic digital transformation and going overseas customers. On digital native customers, recently demand remained soft, primarily due to macroeconomic challenges and regulations. However, on this segment, as I mentioned earlier, we also see large opportunities on generative AI-powered use cases. That would take some time. On digital transformation side, recently we saw longer sales cycle, but there is clearly a lot of unmet demand from traditional enterprises. We expect this segment to generate sustained growth for us. On going overseas customers, the overall growth momentum has been strong. On Agora's side, we see overall positive demand momentum in more developed market, especially on the following verticals. Live streaming e-commerce, creator and finance economy, sports live streaming, future of work. For emerging market, We see pressure on our customers due to macro economic challenges and market competition, such as in South Asia and Southeast Asia. That's from the outlook of the demand side. And generative AI side, we actually have invested in this area for quite some time to help improve the audio video quality. And as I mentioned, large language model can actually help create more use cases in many verticals. But if your question is around reduced cost, I don't think it's going to hugely help us to reduce cost. But in terms of disrupting many business models, like in education, AI-powered tutor can be much cheaper than a real teacher or a virtual celebrity or singer can also be much cheaper or virtual matchmaker the same sense or even in social space. As you can see, a popular character who's busy talking to people in one social platform can be also an AI-powered person. to reduce the platform's cost to maintain and pay a real person who's active on that platform. I think on those, it can hugely disrupt their business.

Thank you.

Right, thank you. Our next question comes from the line of Daily Lee from Bank of America Securities. Please ask your question, Daily.

Hi, thanks for taking my question. I have one question regarding our international, the Agora business. For this quarter, it seems to deliver quite solid growth. I know what's the key drivers for the growth, for example, regarding the volume and the and ASP, and what's the outlook for the volume and ASP going forward? Thank you.

As we explained in previous earnings calls, so the market in China and the market in U.S. and other international markets, at different stage of development in terms of adopting the RT technology. China market is comparatively more mature. In the past 18-24 months, we certainly see US market catching up quite rapidly. For example, the live streaming e-commerce use case. that was already popular in China two years ago, and probably even earlier. But that has really just taken off in the past 12, 18 months in the US. And that has driven a lot of new demand for technology. Also, we have been expanding into more geographic regions, South Asia, Middle East, and even South America. So that explains the stronger cost momentum on the agrarian side. And looking forward, as Tony just explained, in China we do see more macro challenges, the overall macroeconomic environment and also regulation. In terms of the more traditional enterprise customers in China, we also see more budget constraints. So looking at the next few quarters, we are also more optimistic in terms of demand for the Acura side.

Thank you, Jinbozong.

Thank you. Our next question comes from the line of Ethan Zhang from Nomura. Please ask your question, Ethan.

Hi, thank you management for taking my question. This is Bing Duan from Nomura. I have one question regarding the China market as we noted that There are a few changes in the competition landscape, such as AliCloud's plan of spinoff and the recent price cut from the large cloud companies like Alibaba and also some China's telecom operators. So how do we see this impact our business in China such as the ASP and margins and also about our future market to share in China? Thank you.

First of all, the public cloud price adjustment recently does not really to us and giving pressures to us. However, in China, as I mentioned earlier, the overall competition has remained strong for a long time and the demand is actually, especially for domestic and native part, is actually soft. With this situation, we do practically enhance our competitive edge Sometimes we reduce our pricing on certain areas to increase competitive pressure and try to gain more market share. For major public cloud, previously also rolling out similar service in RTE sector. Honestly, there's no one being left. Everyone already released some similar product already. course of past two, three years, many of them have mostly exited the business. They either stopped really selling that product or some already started a partnership with us and available our product in their offerings.

Thank you, Tony.

As a reminder, to ask a question, please press star 1 1 on your telephone.

To ask a question, please press star 1 1.

I am showing no further questions. I'll now turn the conference back to Fiona for closing remarks.

Thank you, operator.

Thank you, everybody, for joining our call today. Again, if there are any further questions, feel free to contact us. And also, the presentation and the remarks of this call will be posted on our website. Thank you, everybody. Thank you.

Thank you.

This concludes today's conference call. Thank you for participating.

You may now disconnect.