Alibaba CTO Zhang Jianfeng Explains Alibaba’s Big Data Strategy and Technology Stack

“Whether it is artificial intelligence or other frontier technologies, high‑quality data, a powerful computing platform, and an efficient algorithm platform are indispensable,” Alibaba Group CTO Zhang Jianfeng said in Seattle, emphasizing that only the combination of these three elements can achieve breakthroughs in machine learning and AI.

On August 6, Alibaba held a technology forum in Seattle attended by nearly 400 local tech professionals; besides architecture, middleware, and search leaders presenting practical insights, CTO Zhang Jianfeng also shared Alibaba’s technology strategy for the first time in the United States.

Zhang Jianfeng, nicknamed “Xing Dian,” has been with Alibaba for 12 years, witnessing the evolution of Taobao, Tmall, and Juhuasuan into the world’s largest e‑commerce platform. He has led multiple technology teams, served as president of the China Retail Platform Group, and was appointed Group CTO in April; Alibaba CEO Zhang Yong described him as a rare leader with both technical and commercial experience.

After four months of reflection, Zhang chose this occasion to systematically explain Alibaba’s technology layout around three cores: data, computing, and algorithms.

Why Alibaba Is a Big‑Data Company

Alibaba positions itself as a big‑data company because it possesses massive high‑quality data. “The best big‑data companies today are platform‑centric like Facebook and Google, because they have huge amounts of high‑quality data,” Zhang said, adding that Alibaba’s data is not only abundant but also exceptionally valuable.

Alibaba’s data has three clear characteristics: it is generated by users’ purchase behavior, making it more authentic than search‑based data; it is highly structured, with product descriptions on Taobao containing over a hundred dimensions; and it is dense and real‑time, with more than 100 million daily users across wireless and PC platforms.

Combined with Alibaba’s multi‑scenario ecosystem, these data advantages give the company a unique position for big‑data development.

Compute Platform Needs Large‑Scale Data Training

In building its compute platform, Alibaba has made extensive technical innovations thanks to large‑scale data training. Beyond work on the open‑source Hadoop ecosystem such as stream and batch processing, Alibaba has developed two highly efficient self‑built platforms: the offline platform ODPS and the real‑time platform Galaxy, which not only handle Alibaba’s massive daily compute workload but also power Alibaba Cloud services for external customers.

“Only through massive practice can we discover more improvement directions, so Alibaba truly has the opportunity to change compute platform efficiency,” Zhang said.

Zhang also explained why Alibaba foresaw the future of cloud computing seven years ago: “Alibaba has always built platform‑based businesses; if the transaction platform can be shared, why can’t computing power? Therefore we realized earlier than most that computing could become a public utility like water, electricity, or gas.”

Today, Alibaba Cloud is China’s largest cloud‑computing platform, offering a complete suite of IaaS, PaaS, and SaaS services.

Efficient Algorithms Unlock Greater Data Value

Regarding algorithms, Zhang believes they must be tightly coupled with industry scenarios; laboratory research alone cannot produce truly efficient algorithms. Alibaba’s greatest advantage lies in its ability to provide diverse and extremely rich scenarios. The combination of data, compute platforms, and algorithms is a crucial future trend.

A powerful compute platform plus efficient algorithms can further mine data value, maximize data efficiency, and create a positive feedback loop. Cloud computing accelerates data fusion—for example, isolated weather data has limited value, but when combined with agriculture or commerce, it generates a huge synergistic effect. Traditional manufacturing that fully leverages big data can also significantly improve yield rates.

Alibaba is actively promoting cooperation with traffic, meteorology, manufacturing, and other industries to generate greater data value. “We firmly believe big data will someday change every industry, so Alibaba is exploring new applications across many fields,” Zhang said.

Alibaba’s Future Technology Layout

Looking ahead, Zhang highlighted VR/AR, artificial intelligence, and the Internet of Things. He noted that the pace of world change far exceeds imagination, with countless new technologies emerging, and that the future remains uncertain.

“From PC to wireless, the iteration cycle is extremely short; many companies haven’t even reacted before we’ve entered the wireless era, and many enterprises are left behind,” Zhang said when discussing the hot VR/AR trend, describing it as a shift from 2D to 3D space.

Current AI directions are numerous, and a breakthrough breakthrough has not yet been identified. In Zhang’s view, the most likely success will come from those who study consumer trends, data, and large‑scale scenarios, and Alibaba will invest more resources in this area.