ML Systems Engineer

Job Overview

As an ML Systems Engineer focused on ML Acceleration, you will be responsible for optimizing data loading, gradient computation, and communication processes. You will work on enhancing distributed training pipelines using PyTorch Distributed, designing and maintaining high-performance GPU kernels in Triton or CUDA, and building efficient data loading pipelines to maximize training throughput. This role is at the intersection of ML research and high-performance systems.

Responsibilities

• Profile and optimize data loading, gradient computation, and communication.
• Optimize distributed training pipelines with PyTorch Distributed.
• Design and maintain high-performance GPU kernels in Triton or CUDA.
• Build robust data loading pipelines to maximize training throughput.

Requirements & Qualifications

• Bachelor’s, Master’s, or PhD in CS, CE, or related technical discipline.
• Strong proficiency in Python.
• Extensive hands-on experience with PyTorch.
• Experience optimizing model execution during training and inference.
• Strong ML concepts.