Senior Research Scientist (Foundational)

A new category of AI system is emerging

You will help build the foundations of this category ahead of its inevitable global application.

You’ll work in a small, highly technical team alongside people who have already shaped state-of-the-art AI systems at global scale - individuals operating at the very top tier of real-world AI impact.

The work is backed by some of the world’s most respected technology investors, providing both the runway and the ambition to pursue this properly.

What We’re Looking For

We’re looking for deeply technical and creative researchers who thrive on invention - motivated by open-ended problems and excited to define what’s possible next.

This is a senior role for researchers who enjoy working from first principles, exploring ideas that don’t yet have established benchmarks, and collaborating closely with engineers to ensure research translates into real systems.

What You’ll Be Doing

• Pursue original research into the computational foundations of intelligent systems that model, reason about, and act within complex environments
• Formulate and own an independent research agenda: identifying unsolved questions, designing hypotheses, and developing novel algorithms
• Explore new architectures, learning objectives, and training paradigms beyond incremental improvements to existing approaches
• Build experimental systems and prototypes that demonstrate new capabilities, not just theoretical gains
• Collaborate closely with research engineers and systems teams to ensure ideas are testable, scalable, and grounded in real constraints
• Contribute to the scientific culture of the team through mentoring, discussion, and knowledge sharing
• Publish and present research where appropriate, helping shape the broader research landscape

What You'll Need

• PhD in Machine Learning, AI, Robotics, Computer Vision, or a closely related field (or equivalent research experience)
• Senior or staff-level research experience with a deep understanding of Machine Learning and a track record of original contributions to multi-model learning, video or related areas
• Published in respected venues , such as NeurIPS, ICML, ICLR, CVPR, ICCV, ECCV, CoRL, RSS, ICRA, IROS , or equivalent top-tier conferences or journals
• Comfortable working from first principles: defining problems, designing experiments, and validating ideas rigorously
• Motivated by understanding as much as by performance - you care about why systems work, not just that they do
• Able to operate effectively in small, highly autonomous research environments
• You value research that ultimately informs real systems
• Collaborative by nature, with a bias toward research that can ultimately ship

Location: London

Ideally onsite, with flexibility for exceptional candidates based in the UK or Europe.