Description

You will design and build advanced agent harnesses, create rigorous benchmarks for complex tasks, and collaborate with product teams to solve real-world challenges. This role is critical to making Claude a more capable and reliable agent for long-horizon tasks.

Requirements

• Experience developing complex agentic systems using LLMs
• Significant software engineering and ML experience
• Experience prompting and/or building products with language models
• Strong communication skills and interest in collaborative research
• Passion for making powerful AI safe and beneficial

Responsibilities

• Ideate, develop, and compare agent harnesses (memory, context compression, communication architectures)
• Design and implement rigorous quantitative benchmarks for large-scale agentic tasks
• Assist with automated evaluation of Claude models and prompts across training and product lifecycle
• Collaborate with product org to solve challenging problems applying agents to products
• Help create and optimize data mixes for model training to maximize agentic performance