From Abdul Hamid A Rahman
GUIYANG, CHINA, Sept 1 (Bernama) -- A roughly 40-minute drive from Guiyang to the Gui’an New Area reveals a concentration of technology companies and digital infrastructure that may feel somewhat familiar to a Malaysian visitor, as it bears some resemblance to Cyberjaya.
The visit was part of a media programme organised by Chinese news portal People’s Daily Online in conjunction with the China International Big Data Industry Expo 2026, bringing journalists to Gui’an to observe Guizhou’s developing digital and artificial intelligence (AI) industries.
Across three companies visited by journalists, Gui’an demonstrated how Guizhou is developing an integrated AI value chain - from preparing training data and converting information into commercial assets to providing computing power and access to foundation models. The companies are also looking beyond China, with one believing its data-to-AI model could be replicated in Malaysia and another preparing to make its computing services available internationally.
At the upstream end of the chain, Guizhou Duanmu Youcaitong Big Data Co Ltd prepares and annotates the large quantities of data required to train AI systems. Its services cover autonomous driving, embodied AI, intelligent manufacturing, and multilingual audio and video applications.
The company’s general manager, Zhao Shungang, said Duanmu Youcaitong has handled Southeast Asia-related data-annotation work for major Chinese technology companies with overseas platforms. However, Zhao said expanding data-annotation operations internationally was not as straightforward as transferring the company’s existing model into another country. “One of the core issues is data security and data ownership. For internationalisation, this remains a barrier because data policies do not currently allow everything to be interconnected across borders,” he explained.
This means any expansion into markets such as Malaysia would have to consider local data rules, ownership requirements and where the information is processed.
Developing AI talent
Duanmu Youcaitong also works with universities to train students and develop a pipeline of workers for the AI data industry. Zhao named Guizhou Normal University, Guizhou University, Guizhou University of Commerce and Guizhou University of Traditional Chinese Medicine among its education partners.
“In the months before students graduate, during their internship period, we introduce actual order-based tasks into the universities for training,” he said. “We may also bring them to our company for internships. If they perform well, they may remain here to work.”
The model illustrates how the AI data industry can create employment beyond highly specialised research positions, as preparing and inspecting training data also require human input.
If annotated information is the raw material for AI, Guizhou ChinaDataPay Network Technology Co Ltd represents the next stage - transforming data into assets and commercially usable products.
Founded in Gui’an in 2016, ChinaDataPay provides services covering data governance, confirmation of data rights, assetisation and compliant data circulation.
It helps government bodies, state-owned organisations and companies unlock information that might otherwise remain underutilised or isolated within separate data systems.
The company has also established a “Nine-Step Method for Recognising Data Assets on the Balance Sheet.”
ChinaDataPay said it has developed more than 1,800 data products and implemented over 300 application scenarios across sectors including finance, insurance, transportation, logistics, digital government and tourism. Its work includes converting raw information into applications such as identity verification, risk assessment, vehicle-insurance scoring and traffic-flow forecasting.
Nie Biao, rotating general manager of ChinaDataPay’s Local Data Business Division, said the company’s model, encompassing the transformation of data into assets and subsequently into AI applications, could be taken beyond China and adapted for use in Malaysia.
“From data to assets and then to AI, this is a model that we can replicate in Malaysia,” he said. “It can be replicated through a relatively lightweight approach. We believe the whole process can also be applied in Malaysia.”
Such a model could be relevant to Malaysian government agencies and companies seeking to make better use of existing information while ensuring its ownership, governance and circulation comply with local requirements.
Completing the value chain is Guizhou Suanjia Computing Services Co Ltd, which provides the computing power and model access required to develop and operate AI applications. Jointly incubated by the Gui’an Supercomputing Centre and Guizhou University, the company specialises in distributed computing-power scheduling.
The company said it has connected 11 computing centres across eight Chinese provinces and integrated more than 10,000 AI accelerator cards into its network. Its infrastructure serves SMEs, individual developers, universities and researchers, with the company recording nearly 60,000 individual users.
Its Suanqiao application programming interface (API) platform provides access to mainstream Chinese AI models, including DeepSeek, Qwen, GLM, Kimi and MiniMax, through a single API key.
Daily token usage on the platform has grown from tens of millions when it was launched to the billion-token level, the company said, helping lower the technical and financial barriers for smaller businesses and research institutions seeking to use AI.
The company is now preparing to extend its services internationally.
“We are developing international versions of our two products, Suanjia Cloud and the Suanqiao API platform,” said Yang Jiawei, general manager of Guizhou Suanjia Computing Services Co Ltd. “They are expected to be launched by the end of September and will be available to international users.”
Southeast Asian markets
The international platforms could provide a route for customers in Southeast Asia, including Malaysian SMEs, universities, researchers and developers, to access computing capacity and AI models without having to build costly infrastructure of their own.
Yang said university lecturers and students account for about half of the company’s main target customers. Despite the scale of its network, Suanjia operates with a relatively small workforce. “We currently have slightly more than 40 employees. Last year, our output value exceeded 30 million yuan,” he added.
Taken together, the three companies demonstrate the different components needed to build an AI economy: Duanmu Youcaitong supplies annotated training data, ChinaDataPay converts information into governed and commercially usable assets, and Suanjia provides computing power and access to AI models.
For Malaysia, Guizhou’s experience offers more than a view of China’s technological development. It presents potential areas of collaboration, spanning multilingual AI training data, data-asset management and affordable computing services for companies, universities and research institutions.
Yet, amid the drive to advance AI and commercialise data, signs of tradition remain.
At Guizhou Duanmu Youcaitong, Zhao personally prepared and served tea, refilling each visitor’s cup throughout the interview in a gesture of hospitality and respect. It was a small but telling moment: even as Guizhou’s companies seek a place at the forefront of AI, the traditions shaping how people welcome their guests, conduct conversations and build relationships have not been left behind.
-- BERNAMA
BERNAMA provides up-to-date authentic and comprehensive news and information which are disseminated via BERNAMA Wires; www.bernama.com; BERNAMA TV on Astro 502, unifi TV 631 and MYTV 121 channels and BERNAMA Radio on FM93.9 (Klang Valley), FM107.5 (Johor Bahru), FM107.9 (Kota Kinabalu) and FM100.9 (Kuching) frequencies.
Follow us on social media :
Facebook : @bernamaofficial, @bernamatv, @bernamaradio
Twitter : @bernama.com, @BernamaTV, @bernamaradio
Instagram : @bernamaofficial, @bernamatvofficial, @bernamaradioofficial
TikTok : @bernamaofficial
