Founding Research Engineer, AI Safety & Law

Employia G&A Consulting GmbH; California; Sur place; USD 185 000 - 350 000; Plein temps

Résumé du poste

Employia G&A Consulting GmbH in partnership with an AI safety research lab seeks a Founding Research Engineer to advance law-following AI across evaluations, training, and runtime control. You will design experiments, build evaluation environments and training pipelines, and work with legal experts to ground research in real-world practice.

You’ll own projects end to end, from design through publication. Join a high-ownership, research-driven team tackling frontier AI safety with no playbook,

Qualifications

• 1–5 years in applied research or ML engineering.
• Degree in CS/ML or equivalent experience.
• Experience in AI safety, alignment, evaluations, or related research.
• Strong Python and ML stack skills.
• Ability to tackle ambiguous research problems without predefined solutions.
• Experience building or working with evaluations, training environments, post-training pipelines, or AI agents is highly relevant.
• Comfortable in a fast-paced startup with rigorous research.
• Legal training or research in law/computation is a plus.
• Experience training or validating LLM judges is a plus.

Responsabilités

• Design experiments and build evaluation environments.
• Develop training pipelines and validate LLM judges.
• Collaborate with legal experts to ground research in real-world practice.
• Own projects end to end, from research design through publication.

Connaissances

Applied research

ML engineering

Python

Ambiguity management

Startup experience

Formation

CS/ML degree or equivalent

Outils

ML stack

Evaluation tools

Training environments

Description du poste

Employia G&A Consulting GmbH in partnership with an AI safety research lab seeks a Founding Research Engineer to advance law-following AI across evaluations, training, and runtime control. You will design experiments, build evaluation environments and training pipelines, and work with legal experts to ground research in real-world practice.

You’ll own projects end to end, from design through publication. Join a high-ownership, research-driven team tackling frontier AI safety with no playbook,