Interview with AI Expert Tang Yuan: From Art and Mathematics to Open‑Source Leadership and Distributed Machine‑Learning Infrastructure at Ant Financial

In 2020, the first wave of post‑90s Chinese internet engineers, including Tang Yuan, reflected on their rapid rise from late‑night coding to becoming core forces in AI development.

Tang Yuan, born in 1990, spent his early years studying calligraphy and visual arts before switching to mathematics when he moved to the United States for college, where he discovered programming and began a seven‑year journey in algorithms and technology.

During his undergraduate years he co‑founded the startup DataNovo, focusing on patent data mining and recommendation systems, and later worked as a data scientist at Uptake, building predictive models for IoT applications in aviation and energy.

In 2016, the AlphaGo victory sparked a surge of AI interest in China, but resources were scarce. In February 2017 the first Chinese TensorFlow textbook, TensorFlow实战 , authored by Tang Yuan, filled this gap and received strong endorsement from Google engineers.

Since joining Ant Financial in June 2018, Tang has led the development of AI infrastructure and an automated machine‑learning platform, including the open‑source project ElasticDL, and now focuses on scaling AI workloads on Kubernetes clusters.

He is a PMC member of XGBoost and Apache MXNet, a committer for TensorFlow, ElasticDL, and Kubeflow, and the author of several open‑source tools such as ggfortify and metric‑learn, earning Google’s Open Source Peer Bonus.

When asked about his move to Ant Financial, Tang cited curiosity about domestic tech culture and the opportunity to learn from experienced colleagues from Google, Facebook, and others, as well as a desire to step out of his comfort zone.

He emphasizes the importance of starting as a user in open‑source projects, reading source code, contributing fixes, and engaging with the community to build a personal brand and advance technical skills.

Tang also shares personal reflections on balancing a demanding career with family life, crediting his faith, love for technology, and support from his wife and young son as key factors in his continued growth.

He advises developers over 30 to focus on deep technical foundations, consider both coding and management paths, and stay motivated by long‑term passion rather than age‑related career anxieties.