Research Engineer, Domain Scaling

About the role

The Domain Scaling team at Anthropic builds world‑class AI capabilities for high‑impact knowledge work in finance, healthcare, legal and other verticals. As a Research Engineer you will bridge applied research and data engineering, creating reinforcement‑learning environments that enable the Claude model to excel on real‑world tasks.

Key responsibilities

• Own the end‑to‑end data strategy for knowledge‑work verticals, from task sourcing through RL training.
• Manage technical relationships with external data vendors, evaluate data quality, and design reward signals.
• Collaborate with domain experts to build data pipelines and evaluation frameworks.
• Explore novel methods for constructing RL environments for high‑value tasks.
• Develop QA frameworks to detect reward hacking and ensure environment quality.
• Run generalisation experiments to measure the impact of data‑strategy changes on model capabilities.
• Partner with other RL research and product teams to translate capability goals into training environments and evaluations.

Required profile

• Experience fine‑tuning large language models for specific domains or real‑world use cases.
• Background in reinforcement learning, reward design, or training‑data curation for LLMs.
• Comfort managing technical vendor relationships and iterating quickly on feedback.
• Ability to read and analyse datasets to identify issues.
• Strong cross‑functional collaboration skills and passion for making AI useful across industries.

Required skills

• Fine‑tuning large language models
• Reinforcement learning
• Reward design
• Training data curation for LLMs
• Evaluation and benchmark design for LLMs

What we offer

• Annual compensation range of $350,000 – $850,000 USD.