Staff / Principal Research Engineer, AI Safety, Technical Mitigations
Core
Designing and deploying technical safety systems and safeguards for frontier-class AI models integrated with scientific workflows to mitigate risks from scientific superintelligence.
Role type
Staff / Principal Research Engineer, AI Safety, Technical Mitigations
Builds
Bespoke safety strategy, safety systems, refusal classifiers, automated safety-testing/red-teaming systems, and monitoring systems for scientific data analysis and generation pipelines.
Domain
AI Safety, Scientific Superintelligence, Frontier AI Models
Deliverable
production ML models
Required skills
Building safety systems and classifiers for frontier-class problems, designing scalable production systems, setting research directions for post-training and classifier buildouts, leading teams towards engineering goals
Preferred skills
Developing ML for biological or physical sciences, building safeguards for scientific risks, leading teams
Technologies
Frontier-class language models, lab-based automation, scientific data analysis pipelines
Responsibilities
Set build and research strategy for safety systems across scientific pipelines; Conduct initial safeguards experimentation and buildout; Lead safety systems research to iterate systems beyond state-of-the-art; Partner with domain experts and non-safety teams; Contribute to external communications on safety efforts
Seniority
Staff / Principal, hands-on IC with team leadership
