Hey! I’m Pujith Kachana, an Applied Scientist at Wayve, where I work on the Foundation Model team, focusing on spatial intelligence for autonomous driving and pretraining data. My research focuses on embodied intelligence, particularly spatially grounded perception and reasoning, 4D geometric scene understanding, and action-grounded world modeling.

Previously, I completed my MS in Robotics at Carnegie Mellon University, advised by Dr. Ji Zhang and Dr. Wenshan Wang, and was affiliated with the Field Robotics Center and AirLab. I have also worked at Amazon Robotics as an Applied Science Intern and Software Development Co-op.

Before my time at CMU, I obtained my BS in Computer Science from Georgia Tech, where I worked with Dr. Danfei Xu on deformable manipulation and Dr. Ada Gavrilovska on secure visual servoing.

Research Interests

Ultimately, I want robotic agents to be able to understand and reason about the physical world in an inherently embodied, physically-grounded manner, much like humans. My goal is to develop agents that not only understand geometry, motion, and semantics but can also ground and fuse multimodal concepts, such as language, into this physical world. Specific research topics of interest include 4D reconstruction, world modeling for embodied tasks, and visual-language reasoning.