Applied Scientist, JP OPS STAR
Core
Building decision systems for Amazon Japan's transportation operations using generative AI, large-scale optimization, and graph-based learning.
Role type
Applied Scientist (Generative AI & Operations)
Builds
LLM-based agent systems, GPU-optimized model serving pipelines, and graph neural network models for logistics networks.
Domain
Logistics & Transportation Operations, Generative AI, Large-Scale Optimization
Deliverable
production ML models
Required skills
LLM-based agent system design, graph neural network modeling, GPU-optimized model training and inference, large-scale machine learning system debugging, experimental design and validation
Preferred skills
Patents or publications at top-tier conferences, experience deploying LLMs on AI acceleration hardware (Neuron, TPU)
Technologies
LLMs, Graph Neural Networks, GPU, Java, C++, Python
Responsibilities
Design agentic architectures with structured evaluation frameworks; develop pipelines for large model inference and fine-tuning; build graph neural network models for logistics topology; benchmark latency, throughput, and cost trade-offs; own the scientific lifecycle from formulation to deployment; collaborate with operations stakeholders to validate model outputs.
Seniority
Mid-Senior, hands-on IC