Staff ML Engineer – AWS Trainium, SageMaker

This is a fully remote position, open to applicants in Canada.

📋 Description

• Train and manage models on Amazon SageMaker utilizing AWS Trainium as the primary computational resource.

• Develop and refine PyTorch training code, with a focus on understanding NeuronCore architecture, compiler behavior, and the trade-offs between memory and throughput.

• Identify and resolve hardware-specific training issues, differentiating between data or code-related problems and those linked to compiler or device-level challenges.

• Convert Trainium job requests into effective, cost-efficient end-to-end training pipelines.

• Optimize distributed training sessions for enhanced throughput and cost efficiency on SageMaker infrastructure.

• Collaborate closely with clients and internal engineering teams to define and deliver production training workloads.

⛳️ Requirements

• Extensive hands-on experience with PyTorch, preferably involving distributed or multi-device training.

• Practical experience with Amazon SageMaker for both training and inference tasks.

• Ability to work closely with the hardware layer, including device-specific compilation and debugging at the accelerator level.

• Experience with AWS Trainium or Inferentia (Neuron SDK) is highly advantageous.

• Profound PyTorch expertise with a proven ability to quickly adapt to new hardware targets if prior Trainium or Inferentia experience is lacking.

• Strong foundational skills in Python programming.

• Comfortable operating in a client-facing, production engineering setting.

🏝️ Benefits

• Competitive salary and performance-based bonuses.

• Opportunities for professional development and continuous learning.

• Access to cutting-edge technology and tools.

• Collaborative and innovative work environment.