Eddie Wu吴泳铭
Alibaba Group
CEO
ChinaEddie Wu, male, Chinese national, born in 1975, a native of Hangzhou, Zhejiang, began his career in 1996, holds a bachelor's degree in computer science. He is a director and CEO of Alibaba Group and chairman and CEO of Alibaba Cloud Intelligence Group.
- 1992–1996Studied computer science at the College of Information Engineering, Zhejiang University of Technology, earning a bachelor's degree
- 1996–1999Programmer at China Yellow Pages in Hangzhou(During this period: from 1997, participated in website technology development for China's Ministry of Foreign Trade and Economic Cooperation)
- 1999–2007Co-founded Alibaba (one of its eighteen founders); served successively as technical director of Alibaba (China) and chief technology officer of the B2B platform, Taobao, and Alipay
- 2007–2015Founded the Alimama platform and served as its general manager; later led the mobile Taobao business
- 2015–2023Founded Vision Plus Capital and served as founding partner(During this period: 2015–2020, concurrently served as chairman of the board of Alibaba Health; from 2023, served as chairman of Taotian Group)
- 2023–presentDirector and CEO of Alibaba Group, concurrently chairman and CEO of Alibaba Cloud Intelligence Group(During this period: 2023–2024, concurrently served as CEO of Taotian Group)
Wikipedia · Eddie WuCompany
China's largest cloud and open-model (Qwen) distributor.
Public event locations
Aggregates only locations that have appeared in public reports; locations, order and dates may be uncertain. The line and numbers indicate an inferred route only — not real-time positioning or current whereabouts.
Event timeline
Microsoft’s open-source SkillOpt automatically upgrades AI agent skills without touching model weights
Microsoft released SkillOpt, an open-source framework that optimizes AI agent skill files (.md). Traditionally, skill optimization requires manual instruction editing, which is slow and error-prone. SkillOpt treats skill documents as trainable objects, automatically exploring optimal instruction combinations via deep-learning-style optimization, boosting accuracy for models like GPT-5.5 without altering underlying weights.
Researchers say they trained a foundation model from scratch for about $1,500
Researchers at Sapient developed HRM-Text, using a Hierarchical Recurrent Model architecture, training a 1B-parameter foundation model from scratch for about $1,500 with only instruction-response pairs. The model achieves competitive performance with much larger open models, dramatically reducing pretraining costs and enabling organizations with limited resources to train their own reasoning models.