Pingyi Zhou
Founder & Chief Scientist, BitInf
Former Senior Researcher, Huawei Noah's Ark Lab | Ph.D. in Computer Science, Wuhan University
Founder & Chief Scientist, BitInf
Former Senior Researcher, Huawei Noah's Ark Lab | Ph.D. in Computer Science, Wuhan University
I am a researcher, engineer, and entrepreneur working at the intersection of Foundation Models, Spatial Intelligence, and Physical AI.
Today, as the Co-Founder and Chief Scientist of BitInf, I am leading the development of native vectorized 3D foundation models for the Physical AI era. Our goal is to enable AI systems not only to understand language, but also to understand, generate, reason about, and operate within structured three-dimensional physical environments.
Ph.D. research at Wuhan University. Explored how large-scale software knowledge repositories can be leveraged through NLP, ML, and DL techniques. Focused on software tag recommendation and knowledge understanding using search-learning paradigms, CNNs, RNNs, RCNNs, and Hierarchical Attention Networks.
Joined Huawei Noah's Ark Lab (Speech & Semantic Lab) in 2019, working on dialogue system development and research.
Participated in multiple generations of Huawei Pangu Models, including dense and MoE architectures (billions to trillion-scale parameters). Work covered architecture design, pre-training, efficient training systems, inference acceleration, domain adaptation, and industrial deployment.
At BitInf, leading the development of native vectorized 3D foundation models, building Spatial Foundation Models that bridge language, geometry, engineering knowledge, and physical reasoning.
I believe the next major breakthrough in AI will not come from language alone, but from intelligence that can understand and operate within the physical world.
At BitInf, we are building Spatial Foundation Models that connect language, geometry, physics, and action. This is a long-term research journey that sits at the intersection of Foundation Models, Computer Vision, Graphics, CAD, Robotics, and Physical AI.
I am always looking for exceptional students, researchers, and engineers who are excited by difficult problems and ambitious goals. If you are interested in any of the following areas, feel free to reach out —
We welcome research interns, visiting students, open-source contributors, and future collaborators.
The age of Language AI has just begun. The age of Physical AI is still waiting to be built.
Sparkoh is an AI-native platform for Spatial Intelligence and Engineering Design.
We are building a unified environment where users can create, edit, and interact with structured 3D content through natural language, images, and multimodal inputs.
Core capabilities:
Visit sparkoh.aiA family of foundation models for Spatial Intelligence.
A benchmark for evaluating LLM-based parametric CAD generation systems.
Researching structured vectorized representations for next-generation Spatial AI systems.
From Language Intelligence to Spatial Intelligence
The past decade of AI has been dominated by language models. We believe the next decade will be defined by models capable of understanding and interacting with the physical world. At BitInf, we are building a new generation of Spatial Foundation Models that bridge language, geometry, engineering knowledge, and physical reasoning.
Generate structured 3D representations from multimodal inputs
Enable AI systems to understand geometry, topology, constraints, and physical properties of objects and environments
Develop foundation models capable of reasoning over structured 3D worlds and engineering systems
Connect spatial intelligence with robotics, manufacturing systems, and embodied agents capable of interacting with the physical world
Core contributor.
Open to collaborations in academia, industry, and startup ecosystems.