Principle Machine Learning Engineer

Tocaro Blue is a company focused on transforming maritime intelligence with cutting-edge AI and machine learning technologies. They are seeking a Principal Machine Learning Engineer to lead the development of custom ML models for maritime autonomy, specifically in object detection and tracking in dynamic environments.

Responsibilities:

• Invent and refine custom deep learning architectures for Radar and EO/IR imagery, with an emphasis on semantic segmentation and temporal tracking
• Develop multi-stage ML pipelines (context + characteristic models, segmentation + classification) tailored to low-SNR Radar returns
• Train models on proprietary large-scale datasets (millions of Radar samples and camera sequences) with design-of-experiment methods for data collection and annotation
• Optimize and deploy models to resource-constrained edge hardware (CPU-only and ARM64 platforms), including C++ inference layers
• Advance fusion-aware ML models that integrate Radar with EO/IR, AIS, and cartography for robust classification in GPS-denied or cluttered environments
• Collaborate with fusion and autonomy engineers to ensure ML outputs integrate seamlessly into multi-target tracking and SLAM pipelines
• Contribute to ML-Ops workflows: data management, large-scale training, continuous integration of new field data, and automated evaluation pipelines

Requirements:

• Advanced degree (MS/PhD) in Electrical Engineering, Computer Science, Robotics, or related field
• 7+ years applying machine learning and signal processing to real-world dynamic systems (graduate research counts if directly applicable)
• Demonstrated mastery of semantic segmentation and object classification models, ideally applied to non-vision sensor modalities
• Expert-level Python skills with ML frameworks (TensorFlow/Keras, PyTorch, or equivalent)
• Track record of developing ML models beyond standard YOLO-style detectors, particularly for segmentation of noisy or sparse data (Radar, sonar, or medical imaging)
• Strong background in computer vision and temporal modeling (CNNs, transformers, RNNs for sequential sensor data)
• Experience deploying ML to embedded/edge platforms with optimized C++ inference
• Knowledge of marine, automotive, or aerial robotics systems
• Contributions to large-scale ML data pipelines: annotation strategies, dataset balancing, simulation-to-real transfer
• Passion for pushing the boundaries of AI in GPS-denied, cluttered, and low-visibility environments