Remote AI Safety Research Engineer (NLP & LLM)

Underwriters Laboratories Inc.; Evanston (IL); On-site; USD 98,562 - 135,523; Full time

Job summary

UL Research Institutes is seeking an AI Research Engineer for the Digital Safety Research Institute to design, develop, and evaluate AI safety assessments for NLP/LLMs. The role emphasizes remote work and advancing safety science through rigorous research and publications.

The position calls for a PhD with 3+ years post‑doc experience, deep knowledge of NLP/LLM safety, and collaboration with engineering teams to scale tools and methods in safety assessment.

Qualifications

• Ph.D. in computer science with a focus on NLP or LLMs.
• At least 3 years of post‑Ph.D. experience.
• Mastery of NLP and LLM safety and how assessments protect the public.
• Ability to pursue AI assessment research independently and efficiently.
• Experience partnering with software engineering to build scalable tools.
• Forward‑looking mindset for next‑step analytics on outputs.
• Developing relationships within the safety community.

Responsibilities

• Study incident reports involving LLMs/NLP to hypothesize root causes.
• Review literature and open‑source work in NLP/LLM design.
• Plan and build NLP/LLM assessment models and safety practices.
• Apply diagnostic and troubleshooting skills.
• Code or train text analytics models for NLP/LLM assessments.
• Prepare technical docs for technical audiences.
• Publish research results online and in conferences/journals.
• Collaborate with team to ensure accuracy and rigor.

Skills

NLP

LLM safety

Research

Python

Engineering collaboration

Education

PhD in Computer Science (NLP/LLM)

Job description

UL Research Institutes is seeking an AI Research Engineer for the Digital Safety Research Institute to design, develop, and evaluate AI safety assessments for NLP/LLMs. The role emphasizes remote work and advancing safety science through rigorous research and publications.

The position calls for a PhD with 3+ years post‑doc experience, deep knowledge of NLP/LLM safety, and collaboration with engineering teams to scale tools and methods in safety assessment.