Thomson Reuters Labs builds and ships custom models for professional workflows requiring fiduciary-grade AI. Its Thomson LLM family is trained from leading open-weight models using decades of authoritative Thomson Reuters content and extensive subject-matter-expert input. Research engineers build systems to train, evaluate, and run these models, partnering with research scientists to test algorithms, rewards, and evaluations quickly and at scale. Thomson Reuters is hiring Senior- and Lead-level Research Engineers across two areas of emphasis. Large-scale RL training focuses on infrastructure for asynchronous, long-context reinforcement learning of agentic Mixture-of-Experts models above 300 billion parameters that interact with tools and agent harnesses. Evaluation engineering focuses on building, scaling, and optimising evaluation suites and leaderboards for model releases and product integrations. The role builds and operates reliable systems for training, evaluation, and experimentation as models and adoption grow. It involves profiling and optimising cost, latency, and GPU utilisation, debugging numerical, distributed, and performance problems, and working with training and evaluation teams to bring model releases into production. The position also contributes to production readiness, including the ISO 27001 compliance path, and establishes standards for code quality, testing, and reproducibility. At Lead level, the role owns area architecture, plans delivery and staffing with research leads, and mentors engineers.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
Job Type
Full-time
Career Level
Senior