Inside the ACM Distinguished Scientist Award: Alibaba’s Li Fei‑Fei on Database Innovation

On November 8, the Association for Computing Machinery (ACM) announced its 2018 ACM Distinguished Scientists, including Li Fei‑Fei, chief scientist of the Damo Academy database team and head of Alibaba’s Database Business Unit.

ACM is the world’s oldest and largest computing society; it awards the Turing Award. The ACM Distinguished Scientist title is reserved for researchers who have made major contributions and impact in computer science.

Li Fei‑Fei, a world‑leading database systems expert and former tenured professor at the University of Utah, has repeatedly received IEEE and ACM awards for influential papers in the database field.

In 2018 he left multiple offers from multinational companies and top universities to join Alibaba’s Damo Academy, leading research on next‑generation distributed databases, unstructured data management, data security, and intelligent databases, aiming to elevate China’s database technology to the global frontier.

The award also acknowledges the research strength of Damo Academy. Under his leadership, Alibaba’s new distributed database system powers the company’s massive e‑commerce operations, the Double‑11 shopping festival, smart city traffic management, and the national meteorological big‑data platform, which stores minute‑level data from over 60,000 weather stations since 1957.

Alibaba’s CTO and Damo Academy director Zhang Jianfeng noted that Alibaba’s digital economy has accumulated massive application data, prompting the construction of world‑class infrastructure and independent innovation.

Interview

Q: Congratulations on the ACM Distinguished Scientist award. Why did you move from academia to industry, and why Alibaba?

Database systems are core infrastructure. With rapid hardware advances, exploding data volumes, and complex workloads, large‑scale cluster deployment, high‑concurrency testing, and real‑world scenarios are best addressed in industry. Alibaba’s massive business, Double‑11 challenges, and strong commitment to core technology attracted me.

Q: How does leading a team from academic research to engineering for Double‑11 differ?

Academic research focuses on breakthrough, single‑person or small‑team breakthroughs, while large‑scale engineering requires coordinated group planning and execution. Both “reading books” and “reading people” need careful understanding, but books are static while people are dynamic, requiring continuous communication.

Q: Alibaba actively collaborates with academia. What future cooperation directions does Damo Academy plan, and how do you balance academia and industry?

From databases to the broader computer industry, core technology breakthroughs need industry‑academic interaction. Damo Academy has launched the Qing‑orange award, AIR program, joint labs, and visiting scholar initiatives to strengthen this synergy. Balancing involves abstracting industry challenges into research problems and integrating solutions back into products.

Q: As cloud providers enter the database market, which technical directions will Alibaba’s database products explore?

Cloud databases demand high availability, SLA, compatibility, elastic scaling, storage‑compute separation, and security such as encryption and privacy‑preserving queries. Additionally, intelligent scheduling of storage‑compute resources and automated management are key research areas.

Q: For typical enterprise scenarios that are less extreme than Double‑11, how does Alibaba optimize its databases?

Beyond the extreme‑scale X‑DB system, Alibaba offers cloud‑native POLARDB, which emphasizes efficiency, reliability, and compatibility with existing ecosystems, as well as integrated hardware‑software designs and storage‑compute separation for everyday workloads.

Q: At SIGMOD 2018, “Machine Learning & Databases” was a hot topic. How will AI and ML permeate future databases?

Intelligent database kernels and autonomous operation platforms will become core competitive edges. Traditional cost‑based optimization cannot handle the high‑dimensional tuning challenges of modern workloads; machine learning can leverage DBA expertise and runtime data to enable supervised or unsupervised models for kernel optimization and autonomous management.

Q: Outside of work, what hobbies keep you energized?

I enjoy mountain climbing, running, and wilderness trekking in places like Yellowstone and Arches National Parks, as well as reading history and current affairs.

Thank you for the interview, and best wishes for continued breakthroughs in database technology.