Applied Scientist, Agentic AI for Fulfillment Tech

Amazon; Factoria (WA); Sur place; USD 143 000 - 193 000; Plein temps

Avantages offerts par ce poste

Health insurance

401(k) matching

Paid time off

Parental leave

Résumé du poste

Amazon is seeking an Applied Scientist in Amazon Fulfillment Technology to lead agentic AI development for operational decision making. You will train LLMs, develop multi-agent orchestration, and mentor interns while publishing findings.

Based in Bellevue, WA, the role involves building agentic systems deployed in Amazon buildings, evaluating them offline and online, and advancing capabilities in reasoning, tool use, memory, and action execution.

Qualifications

• PhD or MS with 4+ years in CS/CE/ML or related field experience.
• Experience programming in Java, C++, Python or related language.
• 3+ years of building models for business applications.
• Experience in algorithms, data structures, parsing, numerical optimization, data mining, parallel and distributed computing, HPC.

Responsabilités

• Generating training and preference data for specific use cases (reasoning trajectories, tool traces)
• Reward modeling and policy optimization for LLMs: DPO, IPO, RLHF/RLAIF with PPO/GRPO, rejection sampling.
• Supervised fine-tuning on step-by-step trajectories and tool-use traces
• Verbal Reinforcement Learning and Continual Learning
• RL for LLMs, Offline RL and off-policy evaluation
• Agentic memory/state management; episodic and semantic memory; vector search; grounding with RAG.
• Evaluation: developing decision quality metrics, scaling LLM-based evaluations.

Connaissances

Java

C++

Python

Algorithms & Data Structures

Numerical Optimization

Data Mining

Parallel Computing

Distributed Computing

High-Performance Computing

Formation

PhD

MS in CS/CE/ML

Outils

Unix/Linux

Software Development

Description du poste

Amazon is seeking an Applied Scientist in Amazon Fulfillment Technology to lead agentic AI development for operational decision making. You will train LLMs, develop multi-agent orchestration, and mentor interns while publishing findings.

Based in Bellevue, WA, the role involves building agentic systems deployed in Amazon buildings, evaluating them offline and online, and advancing capabilities in reasoning, tool use, memory, and action execution.