Sail Chen

I am a product of two distinct data cultures.

Born in Shanghai and raised in New Jersey, I built my foundation in the rigorous, structured machine of American enterprise analytics. I learned to find patterns in noise and make the invisible visible using Western methodologies.

But the real challenge began when I returned to Shanghai in 2012.

I walked into a digital explosion. While the West had Facebook and Twitter, China had built a dense, walled garden of super-apps: Weibo, WeChat, and eventually Xiaohongshu and Douyin. The ecosystem was faster, more chaotic, and completely opaque to the outside world.

I spent the next decade bridging that gap. I applied American data discipline to Chinese social velocity. I built topic models in R, LGBM models in Python, and manually deconstructed algorithms. And for five of those years, AmorePacific, the world's fourth-largest beauty company, ran its internal China social analytics on systems I built: the dashboards, the data pipelines, the KOL scoring engines.

That accumulated knowledge is now Unrandom.

Unrandom is the instrument I spent ten years wishing I had. Drop in your campaign's posts and the system does the rest: it pulls the engagement data, watches every video frame, transcribes every word, reads thousands of comments in their original Chinese, and benchmarks every KOL against years of real campaign data. Did they show the product, say the message, hit the moment? That used to be the whole job. Now it is table stakes, verified automatically. Then it writes you the report: verdicts, findings, and recommendations, in English and Chinese, at a depth a consulting team would need weeks to match.

Xiaohongshu and Douyin are the first pipelines we connected. TikTok and Instagram are next, and the engine does not care where the data comes from.

It is also, deliberately, independent. Douyin's tools grade Douyin. Xiaohongshu's tools grade Xiaohongshu. Your agency reports on the campaign your agency ran. Everyone in that chain is grading their own homework, except the brand writing the checks. Unrandom answers to the brand alone: one methodology across platforms, one benchmark, and no incentive to make the numbers look better than they are.

We built this for the teams who spend real money on China social and deserve more than vanity dashboards. Marketers who need to know not just what happened, but whether it was good, why, and what to do about it next.

That's what we provide. Every frame. Every comment. Every campaign. Understood.

Wendy was with Unrandom before Unrandom had a name.

It started with a phone call. I had just landed a major project with Amorepacific and needed a UI/UX designer yesterday. I called Wendy to ask for a referral. Her response was classic Wendy: “You don't need a referral. I'll do it.”

No pitch. No negotiation. The most efficient, no-nonsense collaborator I've ever worked with.

Her background reads like a masterclass in scale: designing enterprise solutions at IBM, building platform ecosystems at NetEase Games, and creating digital patient tools at Sanofi and Takeda. She bridges the gap between corporate rigor and startup chaos.

What makes Wendy dangerous is her component-first thinking.

Over four years, she built design systems for us where every button and interaction pattern was modular and reusable. She challenged my instinct to show data for data's sake and taught me that restraint is a feature.

Every moment in Unrandom where you think “this just makes sense”: that's Wendy.

Today, she is a Co-founder and CPO at Brain Health Innovations, building AI-powered medical devices in Singapore. She advises Unrandom on product experience, bringing clinical-grade rigor to our analytics interfaces.

From gaming to pharma to brain health, the thread is the same: she makes complex systems usable for humans. That's exactly what Unrandom needed.