Lead Agentic RL Scientist — AI for Code Tools

Amazon Web Services (AWS); New York (NY); On-site; USD 219,000 - 296,000; Full time

Job summary

Amazon Development Center U.S., Inc. seeks a Principal Applied Scientist to collaborate with leading scientists and engineers on the next generation of AI‑driven development tools.

You will advance RL and post‐training methods for code‑focused LLMs within the Kiro LLM‑Training team, delivering scalable, production‑ready solutions for developers globally. You will publish in top conferences, work with customers to translate research into products, and help redefine the builder experience from

Qualifications

• PhD, or Master’s degree with 10+ years of CS, CE, ML or related field experience.
• 5+ years of hands-on large scale training experience.
• Experience in patents or publications at top-tier peer-reviewed conferences or journals.
• Experience programming in C++, Python or related language.

Responsibilities

• Push the boundaries of reinforcement learning and post‑training methodologies for large language models specialized in code intelligence.
• Invent and implement state‑of‑the‑art agentic RL training recipes that operate at unprecedented Amazon scale.
• Deploy revolutionary products that directly impact the daily workflows of millions of developers worldwide.
• Break new ground in AI and machine learning, challenging what’s possible in intelligent code assistance.
• Publish and present your pioneering work at premier ML conferences (NeurIPS, ICML, ICLR).
• Accelerate innovation by working directly with customers to rapidly transition research breakthroughs into production systems.

Skills

Large-scale training

Publications

Patents

C++

Python

Education

PhD

Master’s degree

Tools

C++

Python

Job description

Amazon Development Center U.S., Inc. seeks a Principal Applied Scientist to collaborate with leading scientists and engineers on the next generation of AI‑driven development tools.

You will advance RL and post‐training methods for code‑focused LLMs within the Kiro LLM‑Training team, delivering scalable, production‑ready solutions for developers globally. You will publish in top conferences, work with customers to translate research into products, and help redefine the builder experience from