MLOps Engineer (JAX, PyTorch, Pallas/Triton)

Role Overview

Cincinnatus LLC is seeking experienced MLOps Engineers with deep expertise in modern machine learning frameworks, including JAX, PyTorch, and kernel-level programming using Pallas or Triton. In this role, you will help build next-generation AI systems by designing, evaluating, and improving MLOps tasks and training data used to develop frontier Large Language Models.

This is a full-time W-2 opportunity where you will collaborate with a leading AI lab to improve AI training infrastructure, distributed ML systems, and model evaluation through high-quality technical tasks and expert reviews.

Key Responsibilities

Develop MLOps Engineering Tasks

• Design challenging MLOps and ML systems tasks covering:

  • Model training infrastructure
  • Distributed machine learning
  • ML pipelines
  • Kernel optimization
  • AI infrastructure

• Write accurate, well-structured reference solutions.

Evaluate Technical Solutions

• Review MLOps engineering tasks and submissions.
• Assess correctness, scalability, efficiency, and implementation quality.
• Provide detailed written technical feedback.

Improve AI Training Data

• Develop evaluation guidelines and detailed scoring rubrics.
• Help research teams improve:
  • Training pipeline design
  • Distributed systems reasoning
  • Kernel-level optimization
  • ML framework performance

Collaborate Across Teams

• Work closely with AI researchers and engineering teams.
• Ensure consistency and technical accuracy across AI training datasets.

Required Qualifications

• 2+ years of professional experience in:

  • MLOps
  • Machine Learning Infrastructure
  • ML Systems Engineering

• Hands-on production experience with:

  • JAX
  • PyTorch

• Experience writing or optimizing custom GPU kernels using:

  • Pallas (JAX)
  • Triton

• Demonstrated career progression.

• Strong written communication skills.

• Ability to work full-time (40 hours per week).

Preferred Qualifications

• Experience with:
  • Distributed machine learning
  • AI infrastructure
  • GPU optimization
  • Kernel programming
  • AI model evaluation
  • Large-scale training systems

Compensation

• $70–$110 USD per hour

Work Arrangement

• Remote (United States)
• Full-time W-2 employment
• 40 hours per week
• No outside engagements permitted