Editor’s Note: There is no shortage of technology news, yet many promising Chinese startups still remain largely unknown to global audiences. At the same time, the constant flow of funding news, product launches, and new technologies makes it harder to see which companies truly deserve attention. DragonLab is TechNode’s effort to close that gap, looking past the noise to discover China’s next wave of innovation—emerging companies, ambitious founders, and breakthrough technologies that have yet to receive global attention.
Eight years after experimenting with blockchain, Black Lake is turning to industrial AI—testing whether factory data and human know-how can translate into faster, more consistent production decisions in China and abroad.
How long should it take to turn a drawing into something a factory can actually make?
A customer sends over a CAD, or computer-aided design, file. Someone still has to determine how to make the part, which machines to use, how long it will take, what it should cost and whether the deadline is realistic.
At many small and midsize factories targeted by Shanghai-based Black Lake Technologies, the company says those decisions have traditionally relied heavily on experienced engineers and veteran workers.
What happens when that knowledge sits in one person’s head? For Black Lake, that question points to the next stage of factory digitization: turning hard-won human know-how into something software can capture, reuse and act on.
Black Lake’s industrial AI agents handle tasks from drawing interpretation and process planning to quoting, scheduling and quality control. Its CAD-to-Process Agent can analyze a drawing in about a minute, with company-reported accuracy above 95%, versus hours of manual work.
The idea sounds distinctly 2026. The underlying problem does not.
Eight years ago, Black Lake was already focused on a problem that still defines its business: turning factory-floor activity into usable data, then using that data to improve decisions.
In 2018, the Shanghai startup was developing a digital manufacturing platform for production data and analytics. After its Series A, it joined Microsoft Accelerator Shanghai, where it worked with Azure technologies to strengthen the system supporting those operations.
At the same time, Black Lake was experimenting with blockchain to improve the traceability and reliability of production records, as well as image recognition for quality control.
Today, the technologies look very different. Black Lake is applying industrial AI agents to tasks such as process planning, quotation, and scheduling.
But from blockchain to AI agents, the underlying manufacturing question has remained remarkably consistent: how can fragmented factory data and human know-how be turned into knowledge that can be shared, reused, and acted on?
Before intelligence comes data
Black Lake was founded in 2016 after founder and CEO Zhou Yuxiang’s earlier attempt to build a manufacturing analytics business encountered a basic obstacle. Zhou recently told Fortune that many factories his previous startup approached had essentially no data to analyze.
He later worked on a factory floor himself, where he observed lengthy production changeovers, information bottlenecks, and decisions that depended heavily on individual experience.
Those observations helped shape Black Lake’s early focus: giving factories a more systematic way to capture production information and coordinate work. By 2019, its pitch already centered on making factories more flexible as consumer demand shifted toward greater product variety.
That emphasis on flexibility remains central to Black Lake’s AI strategy today.
A factory producing millions of identical items has a relatively stable problem. A supplier receiving frequent small orders, new drawings and changing delivery deadlines faces something messier. Machines, materials, labor, processes and available capacity have to be recombined again and again.
Black Lake is increasingly positioning its software earlier in the production process, where manufacturers decide how to respond to an order before work reaches the shop floor. It is a shift taking place across industrial technology, but companies are approaching it from different starting points.
Boston-based Tulip centers on configurable frontline applications, connected equipment and workflows that manufacturers can adapt across sites.
New York-based Augury grew from machine-health and predictive-maintenance software and is extending that foundation with agents for reliability, maintenance and operations.
Siemens, meanwhile, is integrating AI across a broader engineering, automation and industrial-software portfolio.
Black Lake comes from production-management software. Its AI products are built around factory workflows and production constraints, while its commercial strategy has increasingly targeted China’s small and midsize manufacturers alongside larger factories and supply chains.
Can the model travel?
Black Lake now wants to find out whether experience gained in Chinese factories translates abroad.
The company has been serving more than 100 overseas customers across markets including Southeast Asia, Latin America and Eastern Europe. It operates a regional headquarters in Singapore, and its international footprint now extends to more than 10 countries, including Indonesia, Vietnam and Mexico.
A recent example comes from Puebla, Mexico, where a Black Lake digital system was introduced at an automotive plant in 2026.
Production practices vary; factories may already rely on mature enterprise resource planning (ERP), manufacturing execution system (MES), and automation vendors; data rules and integration requirements also differ by market.
The experience accumulated in China’s fragmented supplier ecosystem may be useful abroad, but it cannot simply be assumed to transfer intact.
Black Lake has new money to test that proposition. In April, the company raised nearly RMB 1 billion ($140 million) in Series D financing. The capital would be used primarily to accelerate industrial AI deployment and global expansion. According to the company, it had become profitable and that annual revenue was growing by more than 60%.
The unglamorous side of the AI factory
Industrial AI is often illustrated with humanoid robots, autonomous machines or lights-out factories. But some of the first consequential uses may look far less dramatic.
They may look like a PDF arriving in an inbox; a quotation that has to go out before a competitor responds; a machine that suddenly becomes unavailable; or an experienced engineer looking at a drawing and knowing, almost instinctively, how it should be made.
The challenge here is that industrial decisions are less forgiving than chatbot conversations. A bad sentence can be regenerated; a bad process plan can waste material; a poor quotation can erase margins; and a scheduling mistake can delay an order.
The data underneath those decisions remains a constraint as well. In a 2026 survey conducted with IndustryWeek, Augury found poor data quality to be the leading barrier to AI maturity, cited by 47% of manufacturing leaders surveyed.
That makes Black Lake’s path from 2018 to 2026 a useful example of how industrial AI is moving closer to real production workflows.
Eight years ago, the company was focused on capturing and organizing factory data. Today, it is exploring how AI can use that data to support parts of industrial decision-making.
The technologies have changed—from blockchain to AI agents—but many of the underlying questions remain. Manufacturers still need to turn experience and operational knowledge into information that software can use, determine where automated decisions can be trusted, and adapt those systems to different production environments.
Black Lake’s approach has also extended beyond product design to distribution. Zhou has recently recruited food-delivery riders to help sell its AI products to factories, reflecting the company’s push to reach smaller manufacturers that traditional enterprise software sales teams may not easily cover.
For Black Lake, the next phase may play out both at home and abroad, as the company explores how widely its industrial AI model can be applied across different factories, production environments, and markets.
DragonLab cuts through the noise to find the Chinese startups, founders, and technologies that could shape what comes next.
Have a startup we should know about? Reach us at dragonlab@ovau.ip-ddns.com.