Staff ML Engineer
Core
Train and operate machine learning models on Amazon SageMaker using AWS Trainium compute, optimizing PyTorch training code and building cost-aware end-to-end training pipelines.
Role type
Staff ML Engineer (AWS Trainium/SageMaker)
Builds
Production training workloads and optimized training pipelines for client and internal engineering teams
Domain
Cloud Infrastructure / Machine Learning
Deliverable
production ML models
Required skills
PyTorch, distributed training, Amazon SageMaker, Python, hardware debugging, compiler optimization
Preferred skills
AWS Trainium, Inferentia, Neuron SDK, multi-device training
Technologies
PyTorch, Amazon SageMaker, AWS Trainium, Inferentia, Neuron SDK, Python
Responsibilities
Write and optimize PyTorch training code for Trainium compilation and execution; Diagnose hardware, compiler, device, data, and code-related training issues; Build cost-aware, end-to-end Trainium training pipelines; Tune distributed training runs for throughput and cost on SageMaker infrastructure; Work directly with client and internal engineering teams to deliver production training workloads
