What this is
Meng Jian (formerly Tencent T11, former ByteDance tech lead) runs a US entity called Nextfield Labs LLC, whose product is Kirkify, with a steady stream of payments from US customers. After the account had been running for a while, Mercury — commonly nicknamed "Shuixing Bank" in Chinese; in substance a fintech company whose underlying banking services are custodied by partner banks — required him to prove "genuine in-US business activity."
On September 6, he submitted his first package of 8 pages — US customer payment details, Stripe settlement records, ACH trace numbers — and it was rejected. He expanded it into 10 pages by September 11; rejected again. With the September 21 deadline approaching, he trimmed it back down to 9 pages, listed four specific questions on the PDF's first page asking the reviewer to flag the gaps, the backend status flipped to Resolved, and review passed.
Along the way he used an AI assistant called "Xiaomo" to re-align multiple orders, original receipts, and bank statements — the one and only place AI shows up concretely in his account.
How the industry sees it
This isn't a technical post, but it's been heavily forwarded across the tech community. The reason is blunt: Mercury is one of the first-choice US accounts for Chinese going-global founders, and "business verification" is the invisible wall nearly every small company slams into.
Supporters say Meng Jian broke the process down carefully: mapping the official options, explaining the company-versus-product distinction (Nextfield Labs LLC vs Kirkify), reconciling Stripe settlements against bank deposits line by line, and avoiding piles of unrelated materials. That is directly useful to anyone mid-process or facing a frozen account.
Criticism exists too. One view: three weeks to clear review isn't long — six-month stalls and permanent account closures are more common. The post's closing line, "awaiting the official final conclusion," itself signals this is no guaranteed win. A sharper worry: Mercury's single-channel dependency is too narrow — any tightening of US fintech risk controls leaves a long tail of Chinese SaaS, cross-border e-commerce, and AI tool teams exposed. A further counter-view: three weeks of AI-assisted material cleanup is still manual labor against real legal and bank-audit scrutiny. The post itself admits "we can't confirm which specific change tipped the decision" — experience that may not transfer.
What it means for regular people
For enterprise IT and going-global teams: Going from 0 to 1 on a cross-border business isn't just shipping product. The whole stack — bank accounts, tax, payment rails — has to be operational, and the cycle is usually measured in months, not weeks.
For individual careers: AI tools are landing first in the "uncool" scenarios — compliance and document cleanup. Re-aligning dozens of orders, original receipts, and bank statements line by line is exactly the structured collation large language models are good at.
For the consumer market: Most people don't feel the pressure on this pipeline — but if you're using any overseas SaaS or tool built by a Chinese team, whether it can keep getting paid and keep operating depends on these behind-the-scenes reviews.