SVTR AI Venture Database · AI Venture Meetup: Reverse Going-Global and Re-Growth · November 16, 2025
A driverless mining truck ran for two years in an open-pit coal mine in Xinjiang, and became the world's largest fleet of battery-electric autonomous mining trucks by shipments. The company behind it, Boonray (Boli Technology), has now filed for an IPO on the Hong Kong exchange, backed by BYD. Its story isn't "go overseas to find a new market"; it's packaging an entire system, ground out in China's harshest and most complex mining scenarios, and selling it into Southeast Asia, Australia, and South America (the company said on-site its overseas orders already span 20-plus countries, which we have not independently verified). On November 16, 2025, at the roadshow SVTR held at Shanghai's MoSpace, all eight startups made the same point: China is no longer just the point of departure, but the proving ground for the product.
The China-US asymmetry in the database is the premise of "reverse going-global"
Per SVTR AI Venture Database, among AI companies with a tagged headquarters, 2,894 are in the US and 493 in China, close to 6 to 1. On company density and capital depth, Silicon Valley remains the high ground. Chinese founders can't win on numbers; what they can win is something else: grind a vertical scenario to a depth others can't match, then replicate it globally on supply-chain cost and iteration speed. Boonray is the hardest sample of this logic, with leading players in the open-pit autonomous-driving solutions market taking nearly 80% share.
SVTR founder Allen framed the event with three questions: in a Silicon Valley crowded with funds, whose money should a founder take; beyond money, what real problems should a founder solve; as AI reshapes everything, how should founding and investing themselves evolve. The questions point to one place: shifting attention from "can I raise money" to "can I make money, and replicate it." That is also the real mood shift in the primary market over the past year; after the hot money receded, only teams that can self-fund earn the next check.
What the eight startups share: switching from imagination to the income statement
Not one of the eight pitched model parameters; every one talked about where the cash flow comes from. ASI builds low-altitude 3D-mapping infrastructure feeding spatial intelligence; Digit Master packages hit marketing workflows into digital employees; SLAM XCAM makes a consumer-grade 3D immersive camera for education, travel, and creation; MoYu AI and Yuhe Technology use AI to rewrite pre-sales copy, CRM, and enterprise knowledge delivery. The directions couldn't be more different, but the underlying moves are highly consistent: build in AI capability from day one, drive straight at commercialization, and use asset-light fast iteration to cut the cost of trial and error.
A year ago, most of these teams were still pitching "how strong our model is." A year later, the same people talk instead about customers, orders, and replication paths. That shift itself is worth recording more than any single project.
So what did "reverse going-global" actually change
Reverse going-global isn't retreat, and it isn't staying home; it flips the order of going global. The old default was to build a product for overseas markets first, then come home to raise money and hire; now more teams reverse it, first proving out the product and cost model in China, then moving the battle-tested whole overseas. What Boonray forged in Xinjiang mines, Digit Master on domestic ad battlefields, and Yuhe among local enterprise customers is essentially the same asset: engineering capability and a cost structure forced out under high pressure, still scarce when carried to overseas markets. The variable cross-border industrial capital and USD funds should really watch is concrete: who can turn the "China scenario plus supply chain" edge into sustained overseas orders, rather than a globalization narrative that stays on the slides.
The dozens of institutions in the room, including Hillhouse Capital and Lenovo Capital, are themselves a footnote to this judgment: money is gathering around Chinese AI teams that already make money and can replicate overseas. The real exclusive of this SVTR event isn't which of the eight projects shone brightest, but restoring "reverse going-global" from a slogan into a path investors can actually test.
Join SVTR membership to unlock the deals: The full materials of the eight startups on stage, plus direct lines to their founders, are open to members. Become a member to view project details and connect with founders.