Robotics ML Training Infrastructure Engineer

Eka Robotics; Boston (MA); On-site; USD 140,000 - 190,000; Full time

Job summary

Eka Robotics is seeking a Reinforcement/Machine Learning Infrastructure Engineer to design, implement, and maintain large-scale model training systems powering our robot learning efforts.

You will focus on building a streamlined developer experience and scalable tooling to accelerate research cycles, from prototyping to production training runs, collaborating closely with researchers.

Qualifications

• BS, MS or higher in Computer Science, Computer Engineering, Machine Learning or related technical field.
• Proven track record building ML training infrastructure, internal dev platforms, or scalable systems.
• Hands-on experience with large-scale training using JAX (preferred), PyTorch, or TensorFlow.
• Distributed training on cloud platforms or orchestration systems such as Kubernetes, SLURM, GCP or AWS.
• Strong cross-functional communication and ownership mindset.
• Experience building automated testing pipelines and ML CI/CD workflows.

Responsibilities

• Own Training Infrastructure: design, implement, and maintain large-scale model training systems (orchestration, scheduling, checkpointing, experiment tracking).
• Develop tooling to launch, monitor, debug, and reproduce experiments to improve researcher productivity.
• Scale distributed training pipelines across compute clusters with researchers.
• Manage cloud compute resources efficiently for future scaling.
• Collaborate with researchers to support cutting-edge methods and contribute to training code.

Skills

ML training infra

Distributed systems

Deep learning frameworks

DevOps

Communication

Education

Bachelor's/Master's in CS/CE/ML

Tools

Kubernetes

SLURM

GCP/AWS

Job description

Eka Robotics is seeking a Reinforcement/Machine Learning Infrastructure Engineer to design, implement, and maintain large-scale model training systems powering our robot learning efforts.

You will focus on building a streamlined developer experience and scalable tooling to accelerate research cycles, from prototyping to production training runs, collaborating closely with researchers.