 So is it only in Chinese? No, actually, in a lot of internet giants in China, they all sense in the changes in AI area that conversation will be a big opportunity for access and users. So it's not only for the Chinese market. Is it for English also, or...? Basically, it's for Chinese, because we know Chinese well well. So is it running on Linux, or is it other embedded OS, or what is this? Oh, do OS is actually a pan-OS concept. It's a platform, actually. In consumer electronics, in CE devices, it still runs Linux and Android or free autos. But in the cloud, we have a 2S as a cloud platform. So it's on the cloud, but what is the OS to access it? So what is the run, what hardware does it run on? We provide different SDKs for different hardware platforms like Android SDK, Linux SDK, free autos SDK. All these SDKs connect with the cloud through network, and it collects speech signals and sends it to the cloud. All the processing has to be done in the cloud. So it can run on free autos, like a small Cortex-M series? Yes. So it can be a very cheap smart speaker? Correct. Like very affordable? Yes. It's already for sale in China? Yes. Actually, last year, there are tens of million units has been shipped out with dual OS. Nice. Tens of millions? Yes. What's the price in R&B for the cheapest one? Different form factors, their TV mobile phone and the smart speaker and the smart toy. And for the cheapest one, the cheapest module is a Wi-Fi module. It's just cost a cup of coffee. How much? A cup of coffee. Yeah, a cup of coffee. Like 39 R&B? 49? Even cheaper. Even cheaper? Yeah, 20. 20-something R&B, just a Wi-Fi module? Wi-Fi module, and it's a good microphone inside? Yes. And you can speak to it? Yeah, and you can speak to it, and it collects the signal and sends everything into the cloud, and the cloud help you to understand. What's the trigger word? No, there's no trigger word, just push to talk. Push to talk? Yes. But in general, you have a trigger word on the other smart speaker? A trigger word doesn't work for every scenario. For example, in a smart toy, a child wants to hold a smart toy instead of to wake it up through a trigger word. But what I would like to see in the smart voice assistance is to be able to choose the trigger word. I want, for example, my house, my home to have, I choose the name for it. Is that maybe possible? Have you heard about this? I think it depends on the scenario you are facing. If the device is 3 meters or 5 meters away, it's better for you to wake it up. But I can give you one example. When you're holding a remote controller of TV, definitely there is no need to say a trigger word. You just push the button of the remote controller and say, I want to see somebody's movie and then the movie pops up. So there is no trigger word in this scenario. And it gets improved all the time. When more people use it, it gets better? Yes, correct. And here at the LIMENO Connect, are you talking about the server ecosystem that you might be using to enable this cloud service? Or are you talking about the client hardware open source? So basically we have two platforms for different developers. One is DITP, which is for Intelligent Device Maker. This platform helps those device makers to build a conversational devices. And the other platform is DDP, which is for those application developers. They can build navigation applications or consultant applications. All these applications are based conversation. All right. So what do you think about the LIMENO Connect? You have some good conversations with the other attendees here? Yes. Yeah? Cool, all right.