Senior AI/ML researcher to define ML strategy for a healthcare foundation model and inference system, focusing on frontier models, proprietary domain models, and monetizable features under regulatory constraints. Requires expertise in training/adapting large models, distributed training, RLHF/DPO, retrieval, evaluation, and ML systems engineering, with experience in high-stakes/regulated domains.

Senior Software Engineer role focused on developing, enabling, and performance tuning distributed training solutions for large-scale ML models (LLMs, Stable Diffusion, ViT) on AWS Neuron accelerators using PyTorch. The role involves building distributed training support into PyTorch, the Neuron compiler, and runtime stacks, with a focus on strategies like FSDP, PP, and Context parallel. Experience with post-training strategies is a plus.

Senior Applied Scientist role focused on developing and implementing novel foundation models for logistics, combining multimodal data (image, video, geospatial) and large-scale training/inference. The role involves guiding technical direction, mentoring, and collaborating across science and engineering teams to deploy these models for various Amazon delivery use cases.

Principal Applied Scientist role focused on developing and implementing novel foundation models for Amazon's delivery logistics. The role involves designing deep learning architectures, training models on vast datasets, and ensuring production-level performance for multimodal data (image, video, geospatial). It emphasizes scientific leadership, collaboration, and mentoring, with a strong focus on both research and engineering aspects of foundation model development and deployment at scale.

Lead the development of breakthrough foundation models for robotics, focusing on perception, manipulation, and interaction with the world. This role involves hands-on research, algorithm design, and scaling models for real-world deployment at Amazon scale, with a focus on multi-modal and efficient architectures.

Applied Scientist role focused on developing and implementing novel foundation models for logistics, involving multimodal data, training at scale, and inference. The role spans from data preparation to model training, evaluation, and inference, with a focus on production environments.

Software Engineer role focused on building and tuning distributed training solutions for AWS Inferentia and Trainium accelerators, specifically for large language models and other ML model families. The role involves working with PyTorch, Jax, XLA, and the Neuron compiler/runtime to maximize performance and efficiency on AWS Trainium.