Xiang Deng is Co-Founder at NeoCognition, an AI agent lab developing self-learning agents that specialize like human experts. NeoCognition emerged from stealth in April 2026 with a $40 million seed round co-led by Cambium Capital and Walden Catalyst Ventures.

Xiang earned his PhD from the Ohio State University NLP Group, advised by Dr. Huan Sun and working closely with co-founder Dr. Yu Su. He previously worked with Microsoft Research, Amazon, Google, and Scale AI as a senior research scientist.

Xiang's research focuses on AI agents, coding with large language models, and NLP with emphasis on utilizing knowledge from heterogeneous sources. His work contributed to seminal projects like Mind2Web, MMMU, and SeeAct that helped lay the foundation of the modern AI agent field.

NeoCognition is building agents that continuously learn to become domain experts by constructing world models, targeting enterprises and SaaS companies. The company has 15 employees, most holding PhDs, and is backed by Vista Equity Partners and notable angels including Intel CEO Lip-Bu Tan.

Research from the founding team is widely used in frontier large language models from OpenAI, Anthropic, and Google. NeoCognition aims to democratize access to expertise by developing agents that achieve expert-level reliability and efficiency for high-stakes applications.