Senior ML Engineer - AI/ML on Custom Accelerators

Amazon Inc.; Seattle (WA); Hybrid; USD 168,000 - 227,000; Full time

Benefits offered by this job

RSUs

Sign-on bonus

Health insurance

401(k) matching

Job summary

Amazon seeks a Sr. Software Engineer- AI/ML for AWS Neuron in Seattle. You will design, optimize, and deploy ML models on custom accelerators, working across ML frameworks, compilers, and hardware with a strong emphasis on performance and scalability.

You’ll collaborate with cross-functional teams and customers to push state-of-the-art AI workloads. Role focuses on high-performance kernels, system-level optimizations, and end-to-end ML lifecycle across distributed architectures.

Qualifications

• Bachelor’s degree and 5+ years of professional software development experience.
• Strong Python and/or C++ programming skills.
• Experience designing or architecting large-scale systems and leading design efforts.

Responsibilities

• Design, develop, and optimize ML models on custom AI accelerators.
• Participate in ML system lifecycle including distributed architectures, profiling, and deployment.
• Build infrastructure to onboard multiple models with diverse architectures.
• Design and implement high-performance kernels and features for ML ops on Neuron hardware.
• Collaborate with customers and teams to optimize models for AWS accelerators.

Skills

Python programming

C++ programming

Performance profiling

System design leadership

Debugging

Education

Bachelor's degree

Master's degree

Tools

CUDA

Triton

PyTorch

NCCL

Job description

Amazon seeks a Sr. Software Engineer- AI/ML for AWS Neuron in Seattle. You will design, optimize, and deploy ML models on custom accelerators, working across ML frameworks, compilers, and hardware with a strong emphasis on performance and scalability.

You’ll collaborate with cross-functional teams and customers to push state-of-the-art AI workloads. Role focuses on high-performance kernels, system-level optimizations, and end-to-end ML lifecycle across distributed architectures.