Software Engineer I - AI/ML, AWS Neuron Distributed Training

Annapurna Labs (U.S.) Inc. is seeking a Software Engineer I with a focus on AI/ML distributed training. The role centers on optimizing large-scale models for AWS Trainium, implementing mixed-precision strategies, and extending training frameworks within the Neuron ecosystem, in collaboration with hardware and AWS teams. This is an onsite position based in Cupertino, CA, offering a compensation range of USD 127,100 to 185,000 per year.

Responsibilities

• Contribute to designing and implementing distributed training solutions for large-scale ML models running on Trainium instances.
• Extend and optimize popular distributed training frameworks within the Neuron ecosystem, including FSDP, torchtitan, and Hugging Face libraries.
• Develop and optimize mixed-precision and low-precision training techniques, covering BF16, FP8, and emerging formats to boost throughput while preserving accuracy and convergence quality.
• Implement precision-aware training strategies, loss scaling, and careful gradient management to ensure stability across reduced-precision formats.
• Profile, analyze, and tune end-to-end training pipelines to achieve optimal performance on Trainium hardware.
• Collaborate with hardware, compiler, and runtime teams to understand system constraints and unlock new capabilities.
• Partner with AWS solution architects and customers to support the deployment and optimization of training workloads at scale.

Requirements

• Bachelor's degree or higher in computer science, computer engineering, or a related field (or equivalent bachelor's degree).
• 1+ years of programming experience in at least one software language, including academic projects, internships, or research.
• Experience with software development practices such as code reviews, source control, testing, and build processes.
• Experience with machine learning concepts and at least one ML framework (PyTorch, JAX, or TensorFlow).

Technologies

• FSDP
• torchtitan
• Hugging Face libraries
• PyTorch
• JAX
• TensorFlow
• Trainium
• AWS Neuron
• BF16
• FP8

Benefits

• Health insurance
• 401(k) matching
• Paid time off
• Parental leave
• Sign-on payments
• Restricted stock units (RSUs)
• Flexible Spending Accounts
• Employee Assistance Program (EAP)
• Mental Health Support
• Adoption and Surrogacy Reimbursement