Ruoming Pang

Ruoming Pang is an artificial intelligence researcher and engineer who, as of February 2026, is a member of technical staff and AI researcher at OpenAI, working on foundation models and superintelligence research.^[4][5] He previously served as an AI researcher in Meta’s Superintelligence Labs overseeing AI infrastructure for new-generation models, led Apple’s Apple Foundation Models team behind Apple Intelligence features such as Genmoji, on-device summarization, and updates to Siri and personalization, and spent about 15 years at Google working on large-scale machine learning systems, speech technologies, and global access control infrastructure.^[3][4][1][5]

Education

Ruoming Pang earned a Ph.D. in Computer Science from Princeton University (2000–2006). He previously completed a Master’s degree in Computer Science at the University of Southern California (1998–2000) and obtained a Bachelor’s degree in Computer Science from Shanghai Jiao Tong University (1995–1998).^[1][2][5]

Career

OpenAI

In February 2026, OpenAI hired Ruoming Pang as a member of technical staff and AI researcher after pursuing him for several months amid competition for senior AI researchers, according to reporting by The Information summarized by Reuters and other outlets.^[4][5][6][3]

Meta

From July 2025 to February 2026, Pang worked as an AI researcher in Meta’s Superintelligence Labs, where he oversaw the artificial intelligence infrastructure for new-generation models being developed by the unit.^[3][4][6] He was recruited during Meta’s AI-focused hiring initiative on a compensation package reportedly valued at more than $200 million over several years, reflecting the scale of financial offers used in the competition for senior AI researchers.^[3][4]

Apple

From 2021 to 2025, Pang served as a Senior Distinguished Engineer (also described as a Senior Distinguished Software Engineer) at Apple, where he led the Apple Foundation Models team responsible for the large language models behind Apple Intelligence.^[4][5] He supervised a team of over 100 engineers developing models that power features such as Genmoji, on-device summarization, and updates to Siri and personalization across Apple’s ecosystem, and he was among the authors publicly describing these models in Apple’s 2024–2025 technical reports on Apple Intelligence foundation language models.^[1][3][5]

Google

Ruoming Pang spent approximately 15 years at Google, where he served as a principal software engineer overseeing foundational systems across the company.^[1] During his tenure at Google Brain, he co-led the development of the Babelfish/Lingvo deep learning framework for sequence-to-sequence learning and speech recognition, contributed to Tacotron 2 as a neural text-to-speech model and to the Conformer architecture for speech recognition, and co-authored the EfficientDet and MobileNetV3 neural network families for efficient computer vision.^[5] Between 2012 and 2017, Pang co-founded Google’s global authorization system known as Zanzibar and was its sole technical lead, and he also co-developed a search system built on Bigtable that was subsequently adopted by over 1,000 projects within Google.^[1]