Principal AI Training Platform Lead

Amazon; Cupertino (CA); On-site; USD 208,000 - 282,000; Full time

Job summary

Amazon is hiring a Principal Technical Product Manager to shape and drive training software strategy for Trainium. You will define how researchers and operators train frontier models at scale, from single-node experiments to thousands of nodes, partnering with engineering, research, and customer-facing teams.

The role focuses on distributed training libraries, post-training workflows, and training performance optimization to enable AI/ML workloads efficiently.

Qualifications

• Experience with large-scale model training workflows and distributed training architectures.
• Familiarity with AI/ML training frameworks (JAX, PyTorch) and training libraries interaction.
• Proven ability to drive product strategy and long-term roadmaps across multiple teams.
• Excellent written and verbal communication, including executive-level interaction.

Responsibilities

• Define and execute training product strategy and roadmap from customer requirements with engineering leadership.
• Drive strategy for post-training workflows including RLHF, DPO, and fine-tuning at scale.
• Engage customers with BD, Solutions Architecture, and GTM teams to capture training challenges and requirements.
• Define how Neuron supports the training ecosystem and tools for training workflows on Trainium.
• Lead end-to-end launches for training capabilities with docs, field enablement, and customer communications.

Skills

Large-scale training

Distributed training

JAX/PyTorch familiarity

Roadmap development

Executive communication

Open source collaboration

Education

Bachelor's degree in CS/engineering/analytics/math or equivalent

Job description

Amazon is hiring a Principal Technical Product Manager to shape and drive training software strategy for Trainium. You will define how researchers and operators train frontier models at scale, from single-node experiments to thousands of nodes, partnering with engineering, research, and customer-facing teams.

The role focuses on distributed training libraries, post-training workflows, and training performance optimization to enable AI/ML workloads efficiently.