Founding Member, AI Safety Research & Infrastructure

Sampura; Greater London; Hybrid; GBP 100,000 - 290,000; Full time

Benefits offered by this job

Visa sponsorship

Job summary

Sampura Research is an AI safety non-profit focused on scalable oversight for frontier models. As a founding member of Technical Staff, you will own end-to-end research and engineering problems, build high-quality evaluations, and explore methods to improve judge performance through human–AI collaboration.

You will help design and implement evaluation datasets, leaderboards, and infrastructure while collaborating with the founding team.

Qualifications

• Bachelor's degree or higher in Computer Science, Mathematics, or related field or equivalent experience.
• Proficiency coding in Python or similar languages, with ML tooling (e.g., JAX, PyTorch, Pandas).
• Experience with large language models, human-in-the-loop data annotation systems, or AI evaluation frameworks.
• Willingness to apply AI tools with rigorous human review where needed.
• Track record in exploring open-ended research questions with rigor in academic or industry labs.

Responsibilities

• Design and own end-to-end research and engineering problems within Sampura’s agenda.
• Build diverse, high-quality evaluations for judge protocols from datasets and literature.
• Develop strategies to improve judge performance through hybridization and human-in-the-loop methods.
• Create scalable infrastructure and evaluation pipelines and contribute to benchmark releases.

Skills

Python

JAX

PyTorch

Pandas

Strong research mindset

Open-ended problems

Education

Bachelor's degree or higher in Computer Science, Mathematics, or related field

Tools

JAX

PyTorch

Pandas

Job description

Sampura Research is an AI safety non-profit focused on scalable oversight for frontier models. As a founding member of Technical Staff, you will own end-to-end research and engineering problems, build high-quality evaluations, and explore methods to improve judge performance through human–AI collaboration.

You will help design and implement evaluation datasets, leaderboards, and infrastructure while collaborating with the founding team.