Remote MLOps Engineer: Scalable ML Systems & GPU Kernels

Remote Jobs; United States; Remote; USD 96,000 - 152,000; Full time

Job summary

Mercor is seeking an MLOps Engineer (remote, contract) to guide teams in closing knowledge gaps and enhancing AI model performance across MLOps and training infrastructures. The role requires hands-on experience with JAX or PyTorch at scale and proficiency in GPU kernel optimization using Pallas or Triton.

You will design domain-relevant tasks, provide clear technical feedback, and develop evaluation rubrics for training pipelines while collaborating with experts to ensure data quality.

Qualifications

• 2+ years of dedicated professional experience in ML infrastructure or MLOps.
• Hands-on production experience with JAX and/or PyTorch at scale.
• Experience writing or optimizing custom GPU kernels using Pallas (JAX) or Triton.
• Demonstrable career progression.
• Ability to engage reliably for at least 40 hours/week during weekdays.
• Strong written communication skills and the ability to explain complex technical decisions clearly.

Responsibilities

• Guide research and engineering teams to close knowledge gaps and improve AI model performance in MLOps, training infrastructure, and ML framework-level topics.
• Design challenging, domain-relevant tasks, and write accurate and well-structured solutions to MLOps and ML systems problems.
• Evaluate MLOps tasks and solutions and provide clear, written technical feedback.
• Develop guidelines and detailed rubrics/evaluation frameworks to assess training pipeline design, distributed systems reasoning, and kernel-level optimization across tasks.
• Collaborate with other subject matter experts to ensure consistency and accuracy in training data.

Skills

MLOps

JAX

PyTorch

GPU Kernels

Communication

40h/week

Tools

Pallas/Triton

Job description

Mercor is seeking an MLOps Engineer (remote, contract) to guide teams in closing knowledge gaps and enhancing AI model performance across MLOps and training infrastructures. The role requires hands-on experience with JAX or PyTorch at scale and proficiency in GPU kernel optimization using Pallas or Triton.

You will design domain-relevant tasks, provide clear technical feedback, and develop evaluation rubrics for training pipelines while collaborating with experts to ensure data quality.