Principal Agentic RL Scientist, Code AI Tools

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

Benefits offered by this job

Health insurance

401(k) matching

Paid time off

Parental leave

Job summary

Amazon's AWS Agentic AI team seeks a Principal Applied Scientist to partner with leading scientists and engineers, advancing agentic RL for code-focused models used in Kiro.

You will push RL boundaries at scale, create new training recipes, and bring research into production, collaborating with customers and publishing at NeurIPS, ICML, and ICLR. The role offers impact across developers worldwide and a path to leadership.

Qualifications

• PhD or Master's with 10+ years in CS/ML or related field.
• Minimum 5+ years hands-on large-scale training experience.
• Experience publishing in top-tier conferences/journals.
• Experience programming in C++ and Python.

Responsibilities

• Push the boundaries of reinforcement learning for large language models specialized in code intelligence.
• Invent and implement state-of-the-art agentic RL training recipes at unprecedented scale.
• Deploy revolutionary products that impact developers worldwide.
• Advance AI/ML research and challenge what's possible in intelligent code assistance.
• Publish and present work at premier ML conferences (NeurIPS, ICML, ICLR).
• Accelerate innovation by transitioning research breakthroughs into production with customers.

Skills

C++

Python

Large-scale training

Publications or patents

Leadership / collaboration

Education

PhD or MS +10y CS/ML exp

Job description

Amazon's AWS Agentic AI team seeks a Principal Applied Scientist to partner with leading scientists and engineers, advancing agentic RL for code-focused models used in Kiro.

You will push RL boundaries at scale, create new training recipes, and bring research into production, collaborating with customers and publishing at NeurIPS, ICML, and ICLR. The role offers impact across developers worldwide and a path to leadership.