Anthropic
Research Engineer, Takeoff Intel
About this role
Join Anthropic's Takeoff Intel team to build evaluation and measurement systems that track how AI is accelerating its own development. You'll design and run capability assessments, build data pipelines, and create analysis tools that inform safety research and capability forecasting at scale.
What you'll do
- Design and deploy capability evaluations and measurement instruments for AI R&D tasks
- Develop data pipelines that transform model outputs and telemetry into reliable metrics
- Rapidly prototype new measurement approaches, validate them, and decide what to retain
- Review and supervise AI-generated code as part of standard workflow
- Collaborate with research scientists to identify and prioritize what to measure
- Contribute to internal analyses and public reporting on AI capability trends
What they're looking for
- End-to-end evaluation or data product development
- Rapid prototyping and iterative development
- Large-scale data processing and pipeline design
- Large language model experimentation
- Infrastructure or benchmark design for LLM evaluation
- Python or similar languages for data engineering
- Clear technical communication and cross-functional collaboration
- AI-assisted code review and validation
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Anthropic
Anthropic builds Claude, an AI assistant, and is hiring for engineering roles across infrastructure, data systems, and security that support both AI research operations and the company's internal technology needs. The company seeks infrastructure engineers, systems integrators, data scientists, and security specialists to build production-scale systems for training data pipelines, financial operations, developer productivity measurement, research infrastructure, and server firmware security.
- Website
- anthropic.com
Likely interview questions
- Describe an evaluation or data product you shipped end-to-end—what made it useful and what would you do differently?
- Tell us about a time you worked with messy, large-volume data. How did you balance speed with accuracy?