Lead Research Engineer – AI Evaluation & Benchmarking

Job Description Thomson Reuters Labs is seeking a Lead Research Engineer to lead the design and delivery of production-ready software for artificial intelligence (AI) model evaluation and benchmarking. You will build scalable evaluation infrastructure that enables research and product teams to compare large language models (LLMs), measure quality, cost, and latency, and make informed decisions about model selection and development.

You will own the engineering approach from early prototypes through reliable, maintainable tools, collaborating closely with research scientists, engineering teams, and product partners while helping shape technical direction and engineering delivery.

Key Responsibilities Lead technical delivery and engineering planning:
Partner with program and research leaders to define priorities, break projects into delivery stages, estimate effort and cost, recommend staffing and required skills, identify dependencies, and adapt plans to available resources.

Design and build scalable evaluation infrastructure:
Extend existing shared tools and services, design new components where needed, and develop production-quality Python software, APIs, and reusable evaluation frameworks. Deliver useful initial versions, validate them with users, and improve them through iterative development.

Turn research methods into working software:
Collaborate with scientists to build automated evaluation pipelines connecting model APIs, test datasets, and agreed scoring methods. Implement reliable workflows for comparing LLMs and other AI models using evaluation datasets relevant to products and customer needs.

Ensure reliable, reproducible model comparisons:
Maintain repeatable experiments, separate evaluation data from model development, and make results traceable to evaluation inputs and methods. Measure model quality, operating cost, and latency to support product teams in validating model performance.

Build practical tools for researchers and product teams:
Develop clear user interfaces, documentation, evaluation reports, and diagnostic capabilities. Enable scientists and product teams to run evaluations, understand failures, interpret results, and extend tools without depending on the original developer.

Own production reliability, performance, and responsible data use:
Integrate automated testing, deployment, and monitoring from the beginning. Address performance, scalability, and operating costs, establish service maintenance and support ownership, and work with data specialists to implement access controls, permitted-use rules, and retention requirements.

Collaborate, mentor, and guide engineering decisions:
Partner with foundational research, data engineering, platform engineering, and product teams to align interfaces, responsibilities, and delivery plans. Resolve technical dependencies, mentor engineers, communicate trade-offs and delivery risks, and recommend changes to scope or technical approach when necessary.

Required Qualifications 8-10+ years of relevant software engineering experience, including building and maintaining production systems involving machine learning or complex data.

Strong hands-on Python programming skills, including API design, automated testing, and working with structured and unstructured data.

Technical leadership and end-to-end delivery experience, including architecture decisions, engineering planning, staffing and effort estimation, project execution, and adapting priorities as requirements change.

Experience collaborating with research scientists or machine learning practitioners to translate experiments and research methods into working software.

Experience integrating or testing machine learning models, interpreting evaluation results, and applying…