Tech Stack
Tag name is followed by "@" symbol and proficiency level value.
About proficiency levels:
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
AI @ 3
Claude Code @ 3
Machine Learning @ 3
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
Details
Anthropic is seeking STEM Fellows to work with frontier Claude models and internal tooling on grand challenges across science and engineering. Fellows will work in small pods with researchers to define difficult open problems, build datasets and simulations, evaluate model performance, and identify where Claude falls short of expert judgment.
The four-month program runs from January 12 to May 8, 2027, and is full-time. Some fellowships may be considered for an extension. Fellows are engaged through an employer of record under a fixed-term contract rather than being employed directly by Anthropic.
Responsibilities
- Identify long-horizon open problems where Claude falls short of expert judgment and define what meaningful progress looks like.
- Build comprehensive datasets, simulations, and rigorous, hard-to-game evaluations that test whether models can plan experiments, interpret data, and reason about mechanisms.
- Systematically identify situations where models are confidently wrong.
- Apply Claude to open problems in the Fellow's domain.
- Use test-time compute strategies and integrate the scientific tools and software used in the Fellow's field.
- Participate in a small pod of Fellows working on related problems and meet weekly with Anthropic researchers.
Requirements
- Excitement about the potential for AI to reason about difficult questions where expert judgment matters.
- Hands-on experience using Claude and Claude Code in a day-to-day workflow.
- A PhD in a STEM field and several years of experience, such as postdoctoral or faculty experience.
- Comfort working in a fast-paced, high-agency environment and delivering concrete results over a few months.
- Existing authorization to work in the United States for the duration of the program.
- Prior machine learning experience is helpful but not required.
Program Details
- Access to frontier Claude models and internal evaluation tooling.
- One-week in-person orientation in San Francisco is required.
- The default location is San Francisco; remote engagements within the United States may be considered in certain circumstances.
- Applications are open until October 28, 2026, with offers planned by November 20, 2026.
- Later application stages typically include a take-home exercise and conversations with the research team.
- Anthropic is not currently able to sponsor visas for Fellows.
Compensation
The compensation is $100,000–$167,000 USD for the four-month fellowship, equivalent to $300,000–$500,000 USD annualized, plus other benefits.
More jobs at Anthropic
Developer Relations Manager, Claude Platform
Anthropic · San Francisco, United States, New York City, United States
USD 290,000-435,000 per year
Engineering Manager, Claude Platform
Anthropic · San Francisco, United States, New York City, United States
USD 405,000-485,000 per year
Technical Program Manager, Inference
Anthropic · San Francisco, United States, New York City, United States
USD 290,000-435,000 per year
Staff Forward Deployed Engineer
Anthropic · San Francisco, United States, New York City, United States, Seattle, United States
USD 280,000-320,000 per year
Staff+ Software Engineer, Platform Acceleration
Anthropic · San Francisco, United States, New York City, United States, Seattle, United States
USD 405,000-485,000 per year
Similar jobs
Software Engineer, Labs
Anthropic · San Francisco, United States, New York City, United States, Seattle, United States
USD 320,000-485,000 per year
AI Outcomes Manager
Anthropic · San Francisco, United States, New York City, United States, Seattle, United States
USD 265,000-320,000 per year
Manager, Applied AI Architecture, Enterprise Tech
Anthropic · San Francisco, United States, New York City, United States
USD 0 per year
Applied AI Architect, Public Sector
Anthropic · San Francisco, United States, New York City, United States
USD 240,000-315,000 per year
AI Engineer, GTM Claudification
Anthropic · San Francisco, United States, Seattle, United States
USD 320,000-405,000 per year
Developer Relations
Anthropic · San Francisco, United States, New York City, United States
USD 290,000-435,000 per year
Forward Deployed Engineer - Physical AI Cloud Platform
Nebius · United States, Austin, United States
USD 179,500-224,300 per year
Engineering Manager, Enterprise
Anthropic · San Francisco, United States, New York City, United States
USD 405,000-485,000 per year