Research Engineer
On-siteOxford, England, United Kingdom
Job Summary
Apply and adapt machine-learning and reinforcement-learning research to the product roadmap, translating frontier ideas into shippable capabilities. Design, train, and evaluate models using RLHF and multi-objective policy optimization while balancing accuracy, safety, latency, and cost. Work on LLM post-training including fine-tuning, alignment, and reward modelling, then prototype fast before hardening solutions into production with engineering teams. Define evaluation and auditing protocols for frontier models and engage directly with customers to feed real-world needs back into the research agenda. Help shape the company's research and product vision by bringing a clear point of view on where the field is heading.
Required Qualifications
- A PhD in AI, machine learning or a closely related field, or holding a current postdoctoral research position in Machine Learning
- Strong grounding in Machine Learning and reinforcement learning, including RLHF
- Hands-on experience with multi-objective policy optimisation
- Solid knowledge of transformer architectures and LLM post-training (fine-tuning, alignment, reward modelling)
- Publications at top-tier AI conferences such as ICML, ICLR, NeurIPS, CVPR, etc.
- Experience using HPCs and CUDA for training large-scale models
- The ability to translate research into a product vision and carry it through to delivery
Desired Qualifications
- computer vision experience
- open-source contributions
- applied research that shipped into a product
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.