Staff Machine Learning Engineer, Virtual Collaborator

Job Description

This Staff-level ML Engineer role at Anthropic focuses on training Claude as a best-in-class virtual collaborator. The team designs and implements reinforcement learning environments, tools, and data pipelines that enable Claude to handle complex knowledge work end-to-end across domains like productivity, organizational navigation, and vertical specialties. The role spans finetuning, agent infrastructure, evaluation, and close collaboration with product on agentic workflows.

Responsibilities

• Design and implement reinforcement learning pipelines targeted at virtual-collaborator use cases.
• Build and scale data-creation platforms with domain experts and crowdworkers for open-ended tasks.
• Integrate real organizational data into training environments for Claude.
• Develop rubric-based evaluation systems that maintain quality and avoid reward hacking.
• Train Claude on advanced document understanding and co-creation workflows.
• Partner with product teams to ensure training aligns with shipped features and user workflows.

Benefits