Lead ML Engineer, Robotics Autonomy & Training (Remote)

DaParrot Ltd; United States; Remote; USD 170,000 - 230,000; Full time

Job summary

Serve Robotics is advancing autonomous delivery robotics to move goods more efficiently and safely. The role focuses on building scalable training systems for multimodal data and developing neural architectures that improve autonomy performance across GPU clusters.

You will collaborate with ML scientists and engineers to productionize new models, optimize training pipelines, and monitor experiments to drive continual improvements.

Qualifications

• Master's degree or PhD in CS, Robotics, Electrical Engineering or ML required.
• Minimum 5 years of professional experience with ML models in production.
• Hands-on experience with multi-GPU training and distributed frameworks.
• Strong Python proficiency for ML pipelines, data prep, and training workflows.

Responsibilities

• Design and maintain training systems for petabyte-scale multimodal datasets.
• Identify bottlenecks in data loading, preprocessing, and inter-node communication.
• Develop architectures for autonomy tasks using high-dimensional sensor data.
• Create loss functions and training strategies for complex multimodal inputs.
• Configure large-scale distributed training across machines/GPUs; ensure stability.
• Preprocess, transform, and augment robotics datasets for model training.
• Collaborate with ML scientists to productionize new models and experiments.
• Analyze metrics and logs to drive architectural improvements.
• Develop tools to run experiments and track results efficiently.

Skills

Python

Distributed training

GPU clusters

Neural networks

Multimodal data

Education

Master's or PhD in CS/Robotics/EE/ML

Tools

PyTorch

TensorFlow

Job description

Serve Robotics is advancing autonomous delivery robotics to move goods more efficiently and safely. The role focuses on building scalable training systems for multimodal data and developing neural architectures that improve autonomy performance across GPU clusters.

You will collaborate with ML scientists and engineers to productionize new models, optimize training pipelines, and monitor experiments to drive continual improvements.