Products change the world.
But what truly changes it is technology.
We all came from Big Tech. Then we left.
At those companies, most of our work was enterprise and government projects. Over time one thing became clear: the client who pays and the people who actually use the product are two different groups.
What the client wanted was something to put in a report. Whether users actually used it, whether the problem was actually solved — no one cared. The goal was to pass acceptance, not to build something great — and to do it as cheaply as possible.
Do that long enough, and there's no sense of accomplishment left.
We wanted to build for users — directly.
We figured out what we actually love: building products that go straight to the people who use them — products someone is glad to use, maybe even glad to praise. That praise is the best feedback we've ever had; it reassures us more than any acceptance report could.
Honestly: we're getting older, AI is moving fast, and the anxiety is real. But we've come to believe the best way to fight it is to do something you genuinely love and find meaningful — and make a living along the way.
Building an AI product, for us, happens to be exactly that kind of interesting thing.
What truly changes the world has always been technology.
Products fade, companies disappear — but a problem genuinely solved by technology never comes back. So we don't want to build a shell that just "passes acceptance."
let anyone's video cross the language barrier and be understood by the world.
On a small scale, that's letting a video shot by a Chinese creator make someone on the other side of the planet laugh, or tear up. On a larger scale, it's helping more Chinese brands and Chinese manufacturing reach the world in a language it understands. Language shouldn't be the barrier that keeps good things from getting out.
We insist on getting right what no one else bothers to.
Remember "just pass acceptance"? We refuse.
The pace and rhythm of the dub, whether the lips line up, whether to keep the background music, whether loudness stays even across the track... none of it shows up on a checklist, and users may not even name what's wrong — but they'll feel it's "off," and quietly close the tab.
Erasing subtitles and watermarks from the frame without a trace; swapping a face so naturally you can't find the seam. This is real AI engineering — and what we're proudest of right now.
The seen and the unseen — we take on both. Because we're not building a feature that gets signed off; we're building a product that makes you look twice.
Qiān Yǐng Wàn Yīn — "a thousand images, ten thousand sounds." In every video, what you see is the image; what you hear is the sound. Our name is our promise: to carry both across every language.
If one day your video is truly seen and heard by someone in another language — then we did it right.
We're a group of people who still believe technology can change the world. Maybe a little naive. We intend to stay that way.