ML Research Scientist - Member of Technical Staff
On-siteLondon, England, United Kingdom
Job Summary
Originate and pursue ambitious research directions in machine learning and intelligent systems, designing novel model architectures, learning algorithms, and agentic systems. Run rigorous experiments from diagnostic settings through large-scale validation to refine hypotheses and implement methods. Build high-quality research code, evaluation harnesses, benchmarks, and datasets that accelerate scientific progress across the lab. Collaborate with research engineers to translate promising ideas into reliable frontier-scale implementations and communicate findings through internal reviews, papers, and open-source artifacts. This role sits at the heart of Callosum's technical mission, working with a small team on heterogeneous compute infrastructure to develop the foundations for more capable, efficient, and reliable intelligent systems.
Required Qualifications
- A PhD or equivalent record of exceptional research and technical achievement
- Deep expertise in at least one relevant area of machine learning
- Strong mathematical and experimental fundamentals
- Strong hands-on implementation skills in Python and modern ML frameworks such as PyTorch
- The ability to independently own an ambiguous, long-running research agenda
- The ability to collaborate closely, communicate precisely, and update quickly from evidence
- Experience training, evaluating, or interpreting large-scale foundation models, multimodal systems, or learning agents
- Unusually strong research taste
- High agency
- Intellectual honesty
- A low-ego commitment to collective scientific progress
- Visa sponsorship
- Relocation benefits
- In-person work at the London office
Desired Qualifications
- Demonstrated through influential publications
- Spotlights at top-tier conferences
- Widely used systems or open-source work
- Other significant contributions
- A track record of originating work that materially influenced a research direction, important system, or technical community
- Clear individual ownership of contributions adopted or built upon by strong researchers and practitioners
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.