Xiaohua Zhai is an AI research scientist based in Zürich with nine years of experience building and leading multimodal and inclusive AI research programs at top labs including Meta, OpenAI, and Google DeepMind. He holds a Ph.D. from Peking University and has led teams working on open-weight models (SigLIP, PaliGemma), large-scale multimodal datasets (WebLI), and cultural diversity in models, blending technical leadership with hands-on research. At DeepMind he combined tech lead management with research productivity, and his more recent roles at OpenAI and Meta reflect a steady focus on pushing open, reproducible multimodal systems. His work bridges core model development and dataset curation, and he maintains a publication record documenting this trajectory. A subtle throughline in his career is an emphasis on inclusivity and openness in AI — not just performance improvements but broadening who models serve.
9 years of coding experience
10 years of employment as a software developer
Bachelor's degree Department of Computer Science & Technology, Bachelor's degree Department of Computer Science & Technology at Nanjing University
Doctor of Philosophy (PhD) School of Electronics Engineering and Computer Science, Doctor of Philosophy (PhD) School of Electronics Engineering and Computer Science at Peking University
Official codebase used to develop Vision Transformer, SigLIP, MLP-Mixer, LiT and more.
Contributions:8 reviews, 1 PR, 1 issue in 2 years 2 months
pytorchlitmlpvision-transformervision
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.