Lead AI Safety & Technical Mitigations Engineer

United States Digital Space LLC; Greater London; On-site; GBP 166,000 - 249,000; Full time

Benefits offered by this job

Health insurance

Life insurance

Disability insurance

Flexible time off

Parental leave

Educational assistance

Commuter benefits

Lunch program

Company holidays

Job summary

Lila Sciences is seeking a Technical Mitigations Lead to build safety systems for the safe deployment of our scientific capabilities to the world. You will design safeguards that go beyond the current state-of-the-art across in silico and lab-based workflows.

You will start by leading the initial mitigations build-out and then grow a small team to support this function, partnering with domain experts and cross-functional teams to advance safe AI at scale.

Qualifications

• Track record of building safety systems, classifiers, or post-training for frontier-class problems.
• 4–6+ years working with ML systems in engineering roles.
• Experience building scalable, production-grade ML systems.
• Ability to define research directions for open problems in post-training or safety classifiers.
• Ability to communicate complex technical concepts to non-experts.

Responsibilities

• Lead build-out of safety systems across data analysis, generation, and safety testing.
• Oversee initial mitigations work and scale a small safety team.
• Research beyond state-of-the-art safeguards for in silico and lab-based workflows.
• Collaborate with domain experts and other teams to evaluate and restrict capabilities.
• Contribute to external communications on Lila’s safety efforts.

Skills

Frontier-class ML safety

ML system engineering

Production systems

Research direction setting

Technical communication

Job description

Lila Sciences is seeking a Technical Mitigations Lead to build safety systems for the safe deployment of our scientific capabilities to the world. You will design safeguards that go beyond the current state-of-the-art across in silico and lab-based workflows.

You will start by leading the initial mitigations build-out and then grow a small team to support this function, partnering with domain experts and cross-functional teams to advance safe AI at scale.