(SeaPRwire) –
Enterprise AI is still stuck. Every boardroom pitches digital transformation while their data sits in silos, locked in legacy formats, uncleaned and untrusted. The GPUs arrived. The models arrived. What did not arrive was the plumbing to turn raw enterprise data into production-ready tokens at scale. That gap is exactly where Xunce is betting everything.
Xunce (3317.HK) showed up in Berlin as a strategic partner at the BGD 2026 conference alongside the Invest in Germany Summit. They were sitting across the table from Siemens Healthineers, Airbus, Mercedes-Benz, and Bayer. The company reports overseas revenue hit 11.86 million yuan in the first half of 2026, up 264 percent year-on-year. That growth is real, though it still represents a fraction of total revenue. Over 90 percent of Xunce’s income continues to flow from China. Their TokenOS platform, launched earlier this year, converts heterogeneous enterprise data into standardized, measurable tokens in real time. They have already built a “scenario Token factory” targeting European deployment and claim a customer retention rate above 90 percent across 11 high-barrier industries.
Jensen Huang said at the same summit that every company will eventually operate two factories — one physical, one digital — and that the next wave of AI will merge with robotics, manufacturing, and automotive. His framing is spot-on. But Xunce is pushing a harder point: you do not get there by stacking GPUs. You get there by solving data governance first. TokenOS lets raw data stay inside the customer’s environment while producing secure, traceable tokens that models can actually consume. That distinction matters for EU compliance. It also matters for industrial clients who will never ship operational data to a third-party cloud.
What comes next is telling. Xunce already has agreements with the Shenzhen Data Exchange and the Beijing International Big Data Exchange, forming a north-south domestic coordination. They are now targeting automotive, aviation, energy, and manufacturing deployments in Europe as replicable global samples. The risk is familiar. Chinese AI infrastructure firms have scaled quickly at home, only to stumble when facing unfamiliar regulatory frameworks abroad. TokenOS is positioned correctly on paper, but execution in European industrial settings is a different game entirely. Watch the partnership announcements, not the marketing decks.