Zhong'ang Cai is a dedicated Staff Research Scientist at SenseTime, specializing in spatial intelligence for multimodal foundation models. He leads the development of innovative projects like SenseNova-SI and EASI, focusing on scalable spatial training and comprehensive evaluation of advanced models. As the founder of DLP3D, he has pioneered an open-source framework for real-time autonomous 3D characters powered by large language models. With a Ph.D. from MMLab@NTU, he has made significant contributions to the field, publishing over 40 papers in prestigious venues such as NeurIPS, ICLR, CVPR, ICCV, and ECCV.