SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 88,000 - 144,000 / annual
Lila Sciences is building Scientific Superintelligence to solve major scientific challenges through AI systems that autonomously execute the scientific method. You will lead the AI Safety Data and Evaluation Program, a critical function that generates evidence to inform safety decisions across the organization.
Your primary responsibilities center on program execution and data quality. You will translate scientific risk briefs into executable evaluation and red-teaming campaigns, coordinating both internal scientists and external AI safety experts. You will manage contributor onboarding, task assignment, and training to ensure specialists can contribute effectively and consistently. You will establish and enforce quality controls covering completeness, consistency, provenance, and adherence to rubrics, coordinating calibration, review, and disagreement resolution with scientific owners.
You will source and coordinate external experts alongside internal scientific colleagues, organizing access to relevant expertise within Lila. You will run expert-data collection, annotation, red-teaming and evaluation programs, tracking progress and resolving delivery problems. You will maintain clear records, versioned guidance, and usable datasets, continuously improving workflows based on quality findings and researcher feedback. You will identify questions requiring scientific judgment and escalate them to appropriate specialists, ensuring decisions are reflected in program materials and subsequent work.
This role sits at the intersection of AI safety strategy, research operations, and program management. You will work closely with scientists who frame research questions and interpret their significance, creating the operational conditions for their expertise to have impact. You will also gather targeted feedback from partners in the AI Safety community on Lila's safety approach.
Lila is a well-funded, talent-dense organization building systems for scientific discovery across medicine, materials, and energy. The company combines advanced AI models with proprietary instruments into an operating system for science. You will operate with startup speed while tackling problems of historic importance.
REQUIREMENTS:
- Experience delivering expert-data, research-operations, evaluation or similarly demanding programs involving specialist contributors
- Substantial working understanding of AI safety, including threat models, capability evaluations, red teaming, and safeguards, and how they inform one another
- Ability to interpret safety research briefs, recognize assumptions, understand consequences of data-quality failures, and identify when questions need scientific escalation
- Evidence of applying safety understanding in research, evaluation, or expert-data work, including how it shaped program scoping, execution, or review
- Strong operational judgment and ability to manage dependencies, changing requirements, and parallel workstreams
- Track record of setting measurable quality standards, detecting recurring problems, and improving data-producing processes
- Clear communication, careful documentation, and ability to build effective relationships with researchers, experts, and vendors
- Sufficient technical and scientific fluency to understand research briefs, clarify ambiguity, and work productively with subject-matter experts
BONUS:
- Experience contributing to scientific study design, assessing scientific validity, or developing evaluation rubrics
- Deeper scientific expertise or research experience in biology, chemistry, or materials science
- Data analysis, scripting, annotation platforms, or tools for tracking research workflows and data provenance
