Job Description:
We are looking for a Machine Learning / AI Engineer (Part-Time) to design, train, and optimize the perception models that power ROZOR robots.
In this role, you will work with real-world robotics data to build algorithms that enhance environmental awareness—such as object detection, semantic segmentation, behavior prediction, and multi-modal fusion.
Your work will directly impact how ROZOR robots understand people, obstacles, and dynamic spaces as they operate in hospitals, malls, warehouses, and commercial buildings.
Responsibilities:
Train and optimize deep learning models for:
- Object detection, tracking, and classification
- Semantic segmentation and mapping
- Behavior and motion prediction
- Scene understanding in dynamic indoor environments
- Build data pipelines for dataset creation, annotation, preprocessing, and evaluation
- Work with multi-modal sensor data (RGB, depth, LiDAR, radar)
- Integrate ML models into perception and autonomy stacks on ROZOR robots
- Optimize inference performance on embedded compute (Jetson, ARM, GPU)
- Validate model performance through real-world robot testing
- Collaborate with perception, autonomy, and embedded teams to deploy production-ready ML systems
Preferred Qualifications:
- Bachelor’s/Master’s degree (or current student) in Computer Science, AI, Robotics, or related field
- Strong foundation in machine learning and deep learning
- Proficiency in Python, PyTorch or TensorFlow
- Experience training models in detection, segmentation, tracking, or classification
- Hands-on experience with image or point-cloud datasets
- Familiarity with OpenCV, NumPy, and ML tooling
- Strong debugging and experiment evaluation skills
- Ability to work independently in a fast-moving robotics environment