Agentic AI Scientist, Fulfillment Tech (LLM + RL)

Amazon; Bellevue (WA); Vor Ort; USD 143.000 - 193.000; Vollzeit

Benefits dieser Stelle

Health insurance

401(k) matching

Paid time off

Parental leave

RSUs

Zusammenfassung

Amazon.com Services LLC in Bellevue, WA is seeking an Applied Scientist to lead the development of agentic systems for operational decision making and orchestration. You will build full agentic systems with multi-agent orchestration, tool use, memory and action execution, and train LLMs using rejection sampling, SFT and RL methods.

You will evaluate systems offline and online, mentor interns, and publish findings.

Qualifikationen

• 3+ years of building models for business applications.
• PhD or Master’s with 4+ years in CS/CE/ML.
• Experience programming in Java, C++, Python.
• Experience in algorithms and data structures, parsing, numerical optimization, data mining, parallel and distributed computing, high-performance computing.

Aufgaben

• 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.

Kenntnisse

Java

C++

Python

Algorithms and data structures

Numerical optimization

Ausbildung

PhD or Master’s + 4+ years in CS/CE/ML

Tools

Unix/Linux

Software development

Jobbeschreibung

Amazon.com Services LLC in Bellevue, WA is seeking an Applied Scientist to lead the development of agentic systems for operational decision making and orchestration. You will build full agentic systems with multi-agent orchestration, tool use, memory and action execution, and train LLMs using rejection sampling, SFT and RL methods.

You will evaluate systems offline and online, mentor interns, and publish findings.