Good day, everyone. And of this growth, first, strong demand for our voice-enabled wheel robots in China. In Q3, for the second quarter in a row, revenue from these robots doubled year-over-year. They now make up around 5% of our total revenues, supported by both repeat orders and new wins. As of September 30, 2025, the contract backlog for these robots in China was up 32% from the previous quarter. Since then, the backlog has doubled again, reflecting sustained demand. These trends make us comfortable for a continued strong [ UA ] growth in our revenue from these robots in the fourth quarter. Why is demand rising? First, more customers are open to using robots. And today, wheel robots are the most reliable and cost-effective option for large-scale deployment. But more importantly, product experience is getting better. AgentOS, our voice system powered by AI agents, gives our robots a smarter brain to understand and respond more naturally to people. That improved experience allows us to charge a premium even in a competitive market, but most of our revenue growth came from higher shipment volume. We believe AgentOS not only enhances user experience but also strengthens our leadership in voice-enabled robots. Our voice-enabled wheel robot, which integrated with Google's Gemini 2.5 Flash was recently featured by Google Cloud at its AI Asia Conference. We believe this is a strong sign of endorsement. We are continuously upgrading our AI agent capability and applying it to our products. Looking forward, we think these robots can do even better overseas as we combine third-party genAI and multimodal models with our strength in voice AI and autonomous mobility to drive real-world applications. Second, our robotic arm business is growing steadily supported by three key industry trends. Number one, in manufacturing, collaborative robotic arms are becoming more and more popular because they're smaller, easier to install and more affordable, they're also safer to work with. So they help fill many unmet needs in factories like doing tasks that need flexible movement, careful and precise work or real-time feedback. These tasks now rely on human workers today. Number two, in commercial spaces like coffee shops and smart retail, because of advances in lightweight design and easy programming and building vision of feedback, we are unlocking new use cases. Our team's ability to understand real-world needs and build practical products gives us an edge. Number three, robotic arms are a core part of embodied AI. As global demand for physical AI grows, we believe robotic arms will play a key role in bringing AI into the real world. We strengthened our robotic arm business through an acquisition, demonstrating our strategy of combining organic growth with M&A. This business is a great fit for us. It is already profitable with tens of millions of RMB in annual revenue, most of it from overseas customers. By bringing this company into our group, we've expanded our product line and strengthened our presence in global markets. More importantly, we started testing how to combine our wheel robots with robotic arms to create embodied AI that can handle more complex real-world tasks. It's still in early days, but our solid foundation in both technology and product development puts us in a strong position to grow in this space in the long run. Moving on to AI tools. This is another area where we see long-term potential. We're using AI agents to quickly build a variety of new tools for both PC and mobile, and we're also upgrading some of our existing products with AI features. For example, in one of our legacy products, Duba Antivirus users can now interact with their PCs through natural language to complete tasks like system settings. No need for complex manual steps. And small-scale testing of other tools like meeting summarizers, we've also seen strong user engagement and good willingness to pay. What makes this space exciting is that AI coding apps have greatly reduced the time and cost it takes to build and launch new products. This gives us the flexibility to test many ideas quickly and focus on what works. While we're still in the early stages, we believe our strength in building user-friendly tool-based apps, especially with the help of AI agents, puts us in a good position. And since subscriptions already make up more than 60% of our Internet revenue, we're confident in our ability to monetize future products through the same model. To close, I believe Cheetah has moved beyond the turnaround phase. Looking ahead, our focus is on building long-term value by scaling our AI robot business and capturing the upside of AI-native tools. While we're still early, both segments have real momentum and strong potential to drive growth in the years to come.