S

Founding Member of Technical Staff – AI Safety Research (Evaluation, Methods & Infrastructure)

sampura research • United Kingdom
Visa Sponsorship
This Job is No Longer Active This position is no longer accepting applications

Join Sampura Research as a Founding Member of Technical Staff to shape AI safety research focused on scalable oversight. You’ll design novel judge protocols, build high-quality evaluations, and develop infrastructure for hybrid human/AI judge systems. Own end-to-end research and engineering problems in evaluation, methods, and infrastructure with autonomy and impact.

Key Highlights
Founding role shaping AI safety research with autonomy and ownership over key projects
Focus on hybrid human/AI judge systems for scalable oversight of frontier models
Budget of $1M/quarter for compute and human data, with rapid hiring and growth
Key Responsibilities
Design and implement diverse, high-quality evaluations for judge protocols using existing datasets and novel data creation
Research and develop strategies to improve judge performance through hybridization (e.g., learning-to-defer, human assistance)
Build scalable infrastructure for large-scale inference, evaluation pipelines, and human rating platforms
Synthesize findings into papers, blog posts, and benchmark releases for public dissemination
Lead research and mentorship programs through paid fellowships or volunteer initiatives
Technical Skills Required
Python Machine Learning Large Language Models (LLMs)
Benefits & Perks
Visa sponsorship
Competitive salary (£100,000–£290,000 annually)
On-site location in London
Nice to Have
Experience with reinforcement learning, adversarial fine-tuning, or distillation applied to judges
Leadership in mentoring or team leadership for related research programs
Expertise in model confidence/calibration, interpretability, or human-computer interaction

Job Description

About the Org

Sampura Research is an AI safety research non-profit focused on scalable oversight for frontier models through improved human/AI “judges”.

Judges (LLM, human, or a hybrid) are essential for aligning highly capable models through their use in evaluation and training. Limitations in today's judges can lead to undesirable behavior such as reward hacking in training, or failure to spot misalignment in evaluation. As models become more capable and their data more complex, often beyond what the systems overseeing them can understand, we expect these limitations to grow in severity and to become harder to notice.

Want the full job description? Read the complete details on LinkedIn, the original posting.
Continue on LinkedIn

This is a short excerpt. All rights to the full description belong to its original publisher.

See Jaabz jobs first on Google 1 tap · free · in AI Overviews Jaabz is on your Google Manage Preferred Sources
Keep exploring

Similar Jobs

Explore other opportunities that match your interests

Hidden company

Log in to see location

Premium

Senior Applied ML Engineer (World Models & Synthetic Data Generation)

Machine Learning Full-time Mid-Senior level 2h ago

Log in free to see the details

Hidden company

Log in to see location

Premium

Senior Machine Learning Engineer (Algorithmic Trading & Deep Learning)

Machine Learning Full-time 1d ago

Log in free to see the details

Visas, Work Permits & Legal Insights

Stay informed about visa types, work permits, and immigration updates for international professionals.

Browse all articles

Subscribe to our newsletter

New visa sponsorship, relocation and remote tech jobs, delivered to your inbox.