Robotics Algorithm Engineer

Robotics Algorithm Engineer (Perception Scene Reconstruction & Human-Centric AI)
We are seeking a highly skilled Robotics Algorithm Engineer to join our core R&D team and help build next-generation embodied AI this role you will develop cutting-edge algorithms across visual perception 3D scene understanding robot learning and multi-sensor fusion and deploy them directly on real robotic platforms impacting production-level systems.

What Youll Do

• Develop and deploy algorithms for perception 3D reconstruction visual SLAM/VIO and autonomous scene understanding.

• Build robot learning pipelines including imitation learning reinforcement learning and foundation model fine-tuning.

• Create data-driven models leveraging large-scale egocentric and human-robot datasets.

• Architect robust multi-sensor fusion frameworks integrating vision proprioception and tactile sensing.

• Implement production-level software using ROS/ROS2 C and Python.

• Test benchmark and optimize algorithms on physical robotic systems.

• Collaborate cross-functionally on data generation model evaluation and system-level optimization.

• Contribute to publications patents and internal technical roadmaps.

• Mentor junior engineers and support a culture of innovation.

Minimum Qualifications

• MS or PhD in Robotics CS EE or related field (or equivalent experience).

• 3 years experience in robotics computer vision or ML for embodied systems.

• Strong C & Python skills; experience with ROS/ROS2.

• Solid grounding in 3D geometry optimization and estimation.

• Experience in one or more:

  • Visual SLAM / VIO / multi-view geometry

  • 3D reconstruction / mapping

  • Sensor fusion & state estimation

  • Robot learning (IL RL or foundation model adaptation)

• Previous deployment of algorithms on real robotic hardware.

Preferred Qualifications

• PhD with research in robot learning embodied AI or visual perception.

• Deep knowledge of PyTorch/TensorFlow.

• Experience with imitation learning reinforcement learning and policy optimization.

• Experience with CLIP RT-1 VLA or similar VLM/robotics foundation models.

• Background in 3D human modeling MoCap or human-to-robot demonstrations.

• CUDA multi-threading or distributed training experience.

• Publications in ICRA IROS CoRL RSS CVPR ICCV NeurIPS.