Senior AI/ML Distributed Training Engineer

Amazon

Cupertino (CA)

On-site

USD 193,300 - 261,500

Full time

14 days+

Get more replies from employers

Send a job-specific resume in minutes.

Job summary

Amazon Annapurna Labs is seeking a Sr. Software Engineer for AI/ML distributed training to design and optimize large-scale ML training on Trainium instances. You will enhance distributed frameworks and optimize mixed-precision training, working closely with hardware and runtime teams.

Qualified candidates have 5+ years in software development, leadership experience, and strong PyTorch expertise. On-site in Cupertino, you will collaborate with AWS solution architects and customers to deploy

Qualifications

Bachelor's degree in computer science or equivalent.
5+ years of professional software development experience.
5+ years of programming in at least one software programming language.
5+ years of leading design or architecture of new and existing systems (design patterns, reliability, and scaling).
5+ years of full software development life cycle experience, including coding standards, code reviews, source control, build processes, testing, and operations.
Experience as a mentor, tech lead, or leading an engineering team.
Experience in machine learning, large scale training with LLMs, and expertise in PyTorch.

Responsibilities

Design, implement, and optimize distributed training solutions for large scale ML models running on Trainium instances.
Extend and optimize popular distributed training frameworks including FSDP (Fully-Sharded Data Parallel), torchtitan and Hugging Face libraries for the Neuron ecosystem.
Develop and optimize mixed-precision and low-precision training techniques to maximize training throughput while maintaining model accuracy.
Implement precision-aware training strategies, loss scaling techniques, and careful gradient management to ensure training stability across reduced precision formats.
Profile, analyze, and tune end-to-end training pipelines to achieve optimal performance on Trainium hardware.
Partner with hardware, compiler, and runtime teams to influence system design and unlock new capabilities.
Work directly with AWS solution architects and customers to deploy and optimize training workloads at scale.