Applied Scientist, Engineering

Anthropic•San Francisco, CA

1d•Hybrid

About The Position

Anthropic’s Discovery team is working toward building an AI scientist capable of addressing long-term reasoning challenges and advancing scientific and technological capabilities. The team is expanding work in engineering and applied sciences, with the goal of training Claude to invent, design, and engineer across disciplines and scales, reliably and safely. This work may include manufacturing and experimental techniques, processes, materials, devices, instruments, machines, energy systems, and factories. In this role, you will work at the intersection of frontier AI and an engineering or applied science field. You will develop rigorous methods to measure and improve Claude’s performance on complex engineering and scientific tasks, and collaborate with researchers, engineers, and partners at Anthropic and in industry. The work includes enabling AI systems to participate in research and development and testing Claude’s designs against real-world feedback. You will also advise on high-impact problems, tools, and workflows in your field and help translate domain knowledge into machine-learning objectives. The role has a location-based hybrid policy: staff are currently expected to work from an Anthropic office at least 25% of the time, though some roles may require more. The minimum education is a bachelor’s degree or an equivalent combination of education, training, and experience, in a field relevant to the role. The required years of experience depend on the internal job level. Anthropic sponsors visas for some candidates and will make reasonable efforts to obtain a visa for an offer recipient.

Requirements

Deep expertise and hands-on R&D or industry experience in an engineering or applied science discipline, such as mechanical, electrical, materials, chemical, aerospace, structural, industrial, or manufacturing engineering, or applied physics.
Experience with AI or machine learning, whether evaluating or training AI systems or applying these tools in the relevant field.
A track record of taking work from concept through to a working result, such as a product, process, device, experiment, or tool.
Hands-on proficiency with computational design, analysis, or simulation tools used in the relevant field.
Basic programming proficiency, for example in Python.
Ability to navigate ambiguity and develop solutions in a rapidly evolving research environment.
Ability to work independently while collaborating closely with cross-functional teams, and to explain the relevant field clearly to people outside it.
Commitment to using AI to accelerate engineering and scientific progress while maintaining high ethical standards.
Bachelor’s degree or an equivalent combination of education, training, and experience in a field relevant to the role; the posting notes that relevant expertise may be demonstrated through coursework, training, or professional experience.

Nice To Haves

Five or more years of hands-on R&D or industry experience in an engineering or applied science discipline.
Ph.D. in an engineering or applied science field, or equivalent industry experience.
Industry experience in product design and manufacturing, or in developing engineering or scientific simulation software.
Breadth across more than one discipline or across several stages from design to production.
Familiarity with modern AI model training methods.
Track record of building high-fidelity solvers or simulation engines validated against real-world data and used in practice.
Experience working with manufacturers, vendors, contract research organizations, or academic labs.
Experience managing data pipelines and working with large-scale technical datasets.
Publications, patents, or contributions to open-source scientific or engineering software.

Responsibilities

Design and implement evaluation methodologies for assessing AI model capabilities relevant to engineering and applied science research and applications.
Design realistic tasks and training environments, and curate data from the relevant field.
Develop and validate simulations and computational tools, and calibrate them against real-world data.
Work with industry partners, manufacturers, and labs to build and test Claude’s inventions and designs, and incorporate real-world feedback into evaluation and training.
Identify where Claude’s engineering capabilities could have the greatest impact across industry and research, and consider what is needed for adoption.
Advise the team on which problems, tools, and workflows matter most in the relevant field.
Collaborate with domain experts and partners to establish benchmarks and gather high-quality training data.
Translate engineering domain knowledge into machine-learning objectives.