Remote MLOps Engineer: LLM Systems & GPU Performance

mercor; United States; A distancia; USD 124.000 - 165.000; A tiempo parcial

Descripción de la vacante

Mercor, a leading AI talent partner, seeks an MLOps Engineer for LLM systems on a remote contract. You will tackle GPU kernel design, profiling, and inference serving, shaping ML infrastructure and performance.

The role emphasizes collaboration, rigorous feedback, and delivering robust MLOps practices from code to deployment. The ideal candidate has 2+ years in ML systems, CUDA/Triton experience, and strong communication skills, with exposure to JAX/PyTorch and modern accelerators like

Formación

• 2+ years hands-on experience in ML systems or GPU performance engineering.
• Experience with CUDA or Triton kernel programming or profiling tools Kineto/torch.profiler.
• Production experience with JAX and/or PyTorch.
• Familiarity with accelerators such as A100, H100, or TPU.
• Strong written communication skills.

Responsabilidades

• Design challenging tasks across GPU kernels, performance profiling, debugging, and inference serving.
• Guide research and engineering teams to improve AI model performance on ML systems and training infrastructure.
• Develop guidelines and rubrics covering kernel optimization, profiler outputs, and serving throughput and latency trade-offs.
• Collaborate with subject matter experts to ensure training data consistency and accuracy.

Conocimientos

ML systems experience

GPU performance engineering

PyTorch/JAX experience

written communication

Herramientas

CUDA

Triton

Kineto

torch.profiler

JAX

PyTorch

Descripción del empleo

Mercor, a leading AI talent partner, seeks an MLOps Engineer for LLM systems on a remote contract. You will tackle GPU kernel design, profiling, and inference serving, shaping ML infrastructure and performance.

The role emphasizes collaboration, rigorous feedback, and delivering robust MLOps practices from code to deployment. The ideal candidate has 2+ years in ML systems, CUDA/Triton experience, and strong communication skills, with exposure to JAX/PyTorch and modern accelerators like