AI Safety Researcher - London (Fully Remote)

Role: AI Safety Researcher
Type: Contract (6 Months)
Location: London (Fully Remote)

Payrate: 480 - 530 per day on PAYE
550 - 610 per day on RUPAYE
650 - 725 per day INSIDE IR35 Umbrella

What You'll Do

• Red-Teaming: Lead adversarial campaigns to identify system gaps using automated frameworks and LLM-as-a-judge.
• System Alignment: Use Preference Tuning, automatic prompt optimization, and context engineering to align models with safety policies.
• Data Engineering: Query and manage complex datasets to produce high-quality training and benchmarking data.

Who You Are

• Safety Expert: Proven experience in adversarial testing and model alignment.
• Tech Stack: Strong Python, SQL, and Java.
• AI Specialist: Hands-on experience with LLMs and prompt engineering.
• Academic: MSc/PhD in AI/ML (Safety or Agentic systems focus preferred).

This is an urgent vacancy with a deadline where the hiring manager is shortlisting for an interview immediately. Please apply with a copy of your CV or send it praveen. Com

Randstad Technologies is acting as an Employment Business in relation to this vacancy.