Research Engineer

About the Role

We are seeking a Research Engineer based in India who can bridge frontier machine-learning research with shipped product capabilities. This is a critical role for Oxford Dynamics as we navigate an inflection point, defining our growth and impact in mission-critical environments.

Core Responsibilities:

• Translate academic research (Reinforcement Learning, RLHF, multi-objective policy optimisation) into shippable product capabilities using the AVIS platform.
• Design, train, evaluate, and fine-tune Large Language Models (LLMs), focusing on alignment, reward modelling, and rigorous evaluation for safety and latency.
• Rapidly prototype new ideas and collaborate with engineering teams to harden solutions into production systems used by customers globally.

Key Requirements:

• A PhD in AI/ML or current postdoctoral experience is required.
• Demonstrated expertise in Transformer architectures, reinforcement learning, and LLM post-training techniques.
• Prior publications at top-tier conferences (ICML, ICLR, NeurIPS, CVPR) are highly valued.
• Experience with HPCs and CUDA for large-scale model training is essential.

Nice to Have:

• Computer vision experience and contributions to open-source projects.
• A history of applied research that has been successfully shipped as a product.

Note on Location: While the role is based in India, candidates must be able to align their working pattern with UK office hours (RAL Science & Technology Campus time zone).