Thomson Reuters PhD Research Scientist Intern – Foundational Research, London

Overview

Thomson Reuters is offering a hybrid, full-time PhD Research Scientist internship in its Foundational Research division in London (requisition ID JREQ198484). The division focuses on machine learning research and on advanced algorithms and training techniques for large language models, and it keeps strong links with leading academic institutions. The internship typically lasts four to six months, can be aligned with one or two academic semesters, and has flexible start dates through the year.

Interns take part in designing and coding research projects, running experiments, developing models and turning findings into practical deliverables. They work with researchers and engineers at Thomson Reuters and with academic partners, and may share results through seminars, lectures, conferences, publications, data and models. Research areas include large language model training, continued pretraining and instruction tuning, reinforcement learning and alignment, distributed training, planning and reasoning, agentic pipelines and tool use, retrieval-augmented generation, data-centric machine learning, synthetic and hybrid data generation, benchmark and evaluation design, red-teaming and adversarial testing, and hallucination detection and factuality.

Who can apply

The internship is for PhD students and recent PhD graduates with research experience in machine learning, natural language processing, information retrieval or generative AI.

Requirements

The vacancy highlights research experience in a relevant field, publications at top-tier conferences such as NeurIPS, ICML, ICLR, ACL, EMNLP or NAACL, familiarity with deep-learning frameworks such as PyTorch, JAX or TensorFlow, excellent written and verbal communication, and curiosity in solving research problems. Preferred experience includes state-of-the-art LLM research, influential first-author publications, open-source contributions, strong software or infrastructure engineering skills and training large-scale models on distributed cloud infrastructure.

What you get

Interns receive coaching and mentorship from experienced researchers, access to large datasets and major cloud computing platforms, and competitive compensation. The hybrid model supports office-based collaboration.

How to apply

Applications are submitted through the Thomson Reuters careers platform, with the academic background, research experience, publications and technical skills clearly presented.