Research Scientist (Embodied AI & World Models)

This role conducts cutting-edge research in embodied AI and world models for low-power, edge settings.
You will design, implement, and rigorously evaluate ML experiments, generating ideas and actionable results.
You will develop multimodal models for real-time computer vision and audio/video streams in robotics, augmented reality and virtual reality (AR/VR), and autonomous systems.
You will collaborate with hardware and software engineers to define, build and test hardware-aware AI algorithms and new AI accelerators.
You will prepare and present your work at AI conferences and workshops, contributing to publications and knowledge sharing.
Applicants should hold an advanced degree and be proficient in Python (PyTorch/JAX), with strong ML fundamentals and mathematical foundations; exposure to C++/CUDA and low-power hardware is desirable.