Senior Tensor AI Scientist

About this role

Job Description We are looking for a Senior Scientist to join our Advanced Computing & Modelling team and take a leading role in our research at the intersection of tensor networks and modern machine learning. This is an area of strategic importance for Terra Quantum, with applications across several industries. Your mission is to help grow and strengthen this research and to apply it to demanding real-world problems.

We are looking for someone with a broad background in AI and machine learning and strong mathematical foundations, so they can either bring relevant expertise in this area or pick it up and start contributing quickly. The role reports to the Director of Advanced Technologies and works closely with scientists and engineers across the division. Responsibilities Take a leading role in advancing and scaling up our tensor network research, developing new methods for efficient ML model architectures, training, and optimisation, and pushing them beyond the current state of the art.

Design and train modern deep learning models, with a focus on making large-scale training more efficient. Develop and improve the algorithms at the core of our methods, taking ideas from concept to validated result. Work with teams across the division to apply these methods to challenging real-world Contribute to patents and selected publications, and represent Terra Quantum at leading conferences. Help structure the team's research and mentor junior colleagues, raising both scientific and engineering standards.

Collaborate with engineers across the division to run methods efficiently on modern computing platforms. Qualifications Required Must have more than 5 years of relevant post-PhD experience (or equivalent experience after an MSc) with significant relevant experience in machine learning, applied mathematics, computer science, or a closely related field, with strong mathematical foundations in linear algebra, numerical methods, optimisation, and probability.

Hands-on experience training modern deep learning models such as transformers, including efficient large-scale training and fine-tuning, using frameworks such as PyTorch or JAX. Strong command of optimisation methods for machine learning, including gradient-based and large-scale optimisation, regularisation, and training dynamics and Experience with efficient training and model compression techniques, such as low-rank methods, quantisation, distillation, or sparsity.