Machine Learning Research Engineer, Thomson Reuters

• Bachelor's or Master's degree in Computer Science, Engineering, or a relevant discipline (or equivalent practical experience)
• 3+ years of hands-on experience building ML/NLP/AI systems with strong software engineering practices
• Demonstrated expertise in building production-quality code and data pipelines for ML systems
• Proficiency in modern AI development frameworks including: PyTorch, Jax , HuggingFace Transformers, LLM APIs (litellm etc) and vLLM for building and deploying large-scale AI applications
• Understanding of LLM training methodologies including instruction fine-tuning, preference optimization, and reinforcement learning approaches
• Strong software engineering skills including version control, testing, CI/CD, and code quality practices
• Experience with distributed computing frameworks and large-scale data processing (e.g., Ray, Spark, Dask)
• Excellent communication skills to collaborate with researchers and translate research ideas into robust implementations
• Self-driven attitude with genuine curiosity about ML research developments
• Comfortable working in fast-paced, agile environments, managing the uncertainty and ambiguity of genuinely novel research

Desirable

• We are seeking Machine Learning Research Engineers with the skills and drive to build and conduct experiments with advanced AI systems in an academic environment rich with high-quality data from real-world problems
• Foundational Research is the dedicated core Machine Learning research division of Thomson Reuters. We are focused on research and development, with a particular focus on advanced algorithms and training techniques for Large Language Models (LLMs)
• We are expanding our strong foundation of research capabilities across different areas and are looking for engineers who participate in designing, coding, conducting experiments, and translating findings into concrete deliverables
• LLM Training (Continued pretraining, instruction tuning, reinforcement learning, distributed training, efficient ML techniques)
• Post-training techniques for planning, reasoning & complex workflows (e.g., reasoning models, LLMs + knowledge graphs, test time compute, CoT pipelines, tool use & API calling, etc.)
• Data-centric Machine Learning (Synthetic data, curriculum learning, learned data mixtures, etc.)
• Evaluation (Benchmarking best practices, humans/LLMs as a judge, red teaming/adversarial testing, hallucination detection, etc.)
• We work collaboratively with TR Labs (TR's applied research division), academic partners at world-leading research institutions, and subject matter experts with decades of experience
• We experiment, prototype, test, and deliver ideas in the pursuit of smarter and more valuable models trained on an unprecedented wealth of data and powered by state-of-the-art technical infrastructure. Through our unique institutional experience, we have access to an unprecedented number of subject matter experts involved in data collection, testing and evaluation of trained models
• As an ML Research Engineer, you will play a key part in a diverse global team of experts. We hire world-leading specialists in ML/NLP/GenAI, as well as Engineering, to drive the company's leading internal AI model development
• You will have the opportunity to contribute to our proprietary AI model research & development through rapid prototyping, scalable infrastructure, and production-quality implementations, and to research papers in top tier academic conferences and journals
• Build: You will design and implement robust, scalable systems for training and evaluating large language models
• You'll build data pipelines for data-centric research, training infrastructure for instruction fine-tuning (IFT), Direct Preference Optimization (DPO), and reinforcement learning workflows, evaluation frameworks for comprehensive model assessment, and infrastructure for agentic workflows that enables researchers to iterate quickly and effectively
• Innovate: You will work at the very cutting edge of AI Research at an institution with some of the richest data sources in the world
• Through your work, you will rapidly implement novel research ideas in LLM training, evaluation, agentic systems, and data processing, transforming them into production-ready systems and research publications
• You will contribute to advancing the state-of-the-art in data-centric ML by building tools that help us make the best use of our unprecedented data resources
• Experiment and Develop: You are involved in the entire research & model development lifecycle, brainstorming, coding, testing, and delivering high-quality implementations that support cutting-edge research
• You'll build pipelines for synthetic data generation, automated evaluation systems, training workflows for various fine-tuning approaches, and agent-based workflows that push the boundaries of what's possible with LLMs
• Collaborate: Working on a collaborative global team of research engineers and scientists both within Thomson Reuters and our academic partners at world-leading universities
• You'll work closely with researchers to understand their needs and translate cutting-edge research papers into practical, scalable implementations
• Communicate: Actively engage in sharing technical implementations and best practices with the wider team through code reviews, documentation, technical presentations, and knowledge sharing sessions
• Contribute to internal research discussions and stay current with the latest developments in LLM training, evaluation, agentic AI, and data-centric machine learning