Job Description:

We are looking for a Robotics AI Engineer (Deep Learning for Autonomy) to design, train, and deploy neural models that power high-level autonomy behaviors in ROZOR robots.
You will work on dynamic obstacle prediction, semantic mapping, scene understanding, behavior forecasting, and learning-based navigation.

Your contributions will directly enhance how ROZOR robots interpret human movement, anticipate obstacles, and make navigation decisions inside active indoor environments.

Responsibilities:

• Develop deep learning models for:
  • Dynamic obstacle detection and motion prediction
  • Behavior forecasting in crowded environments
  • Semantic mapping and spatial understanding
  • Learning-based navigation and path optimization
• Work with multi-modal data including RGB, LiDAR, depth, IMU, maps, and trajectories
• Build pipelines for training, evaluation, and validation of autonomy-related neural models
• Integrate learned models into ROS2-based autonomy stacks
• Collaborate with planning, perception, and embedded teams to ensure real-time performance
• Conduct robot field testing to validate learning-based behaviors
• Analyze failure cases and retrain or refine models accordingly
• Support dataset generation, annotation, and ML operations workflows

Preferred Qualifications:

• Bachelor’s/Master’s in AI, Robotics, Computer Science, or related field
• Strong foundation in deep learning, computer vision, or reinforcement learning
• Proficiency with PyTorch or TensorFlow
• Experience working with time-series, trajectory data, or sequential prediction models
• Understanding of robot navigation, SLAM, mapping, or motion planning concepts
• Experience training models for detection, tracking, segmentation, or prediction
• Familiarity with ROS/ROS2 and real-time robotics pipelines
• Strong debugging and analytical skills