Build production AI systems remotely from Canada using deep PyTorch expertise across AWS Trainium and Amazon SageMaker training pipelines. Robots & Pencils seeks a Staff ML Engineer to operate and train models on Trainium-powered SageMaker infrastructure. You will reason across PyTorch code, NeuronCore architecture, compiler behavior, memory, and throughput. The work supports enterprise clients through forward-deployed engineering teams delivering production training workloads. Key Responsibilities:

Train and operate models on Amazon SageMaker with AWS Trainium Write and optimize PyTorch training code for Trainium execution Diagnose hardware, compiler, device, data,



and code issues Translate Trainium requests into cost-aware production pipelines Tune distributed training for SageMaker throughput and cost Requirements:

Solid hands-on PyTorch experience, ideally distributed training Production experience with Amazon SageMaker training or inference Understanding of device-specific compilation and accelerator debugging AWS Trainium or Inferentia Neuron SDK experience is valuable Solid Python fundamentals and client-facing production engineering comfort Apply deep PyTorch, SageMaker, and accelerator expertise to production AI workloads on AWS Trainium for enterprise clients.

📌 Staff ML Engineer Trainium SageMaker (Winnipeg)
🏢 Robots and Pencils
📍 Winnipeg