Principal Applied Scientist- Agentic RL , Kiro
Core
Pioneering next-generation AI-driven development tools by creating efficient coding agents for the Kiro LLM-Training team.
Role type
Principal Applied Scientist (Agentic RL)
Builds
Efficient coding agents and revolutionary AI products for software developers
Domain
Generative AI, Large Language Models (LLMs), Reinforcement Learning, Code Intelligence
Deliverable
production ML models
Required skills
Reinforcement learning, post-training methodologies, large scale training, C++, Python, patenting/publishing at top-tier conferences
Preferred skills
Training or inference of large scale models
Technologies
LLMs, Reinforcement Learning frameworks
Responsibilities
Push boundaries of RL and post-training for code-specialized LLMs, implement agentic RL training recipes at Amazon scale, deploy products impacting developer workflows, publish research at premier ML conferences, transition research breakthroughs into production systems
Seniority
Principal, hands-on IC with strategy & mentorship
