Lead/Senior Research Scientist (Agentic RL)

Thomson Reuters•Eagan, MN

About The Position

Thomson Reuters Labs builds custom models for professional workflows that require fiduciary-grade AI. Its Thomson LLM family is trained on open-weight models using authoritative Thomson Reuters content and subject-matter-expert input. The research team trains and evaluates these models for professional workflows, using continual learning, value alignment, data-centric mid-training, DPO, and reinforcement learning. The evaluation group develops expert-annotated legal and tax benchmarks, agentic systems, and human preference studies, and the team regularly publishes its research. The company is hiring Senior- and Lead-level scientists across agentic reinforcement learning, model training, and model evaluation. The role focuses on developing benchmarks, datasets, rewards, and training signals; designing rigorous experiments; building evaluation capabilities; and applying research findings to production models. The scientist collaborates with Research Engineers, legal, tax, and other domain experts, product teams, and model consumers. The role also includes publishing at leading ML and NLP venues and representing Thomson Reuters Labs in the research community. At Lead level, the scientist owns a research agenda, develops a roadmap for model development, and mentors scientists and engineers. This posting is for proactive recruitment and may be used to fill current openings or future vacancies. The position may be eligible for an annual bonus. The posting closes March 31, 2027. Compensation ranges are $127,400–$236,600 USD for eligible US locations and $100,000–$145,000 CAD in Ontario, Canada; pay within the range depends on factors including knowledge, skills, experience, and internal equity.

Requirements

  • PhD in machine learning, computer science, or a related discipline, or equivalent practical experience.
  • Hands-on LLM post-training and/or evaluation experience, preferably at frontier model development scale.
  • Experience using Python and PyTorch.
  • Research impact demonstrated through publications at top-tier venues (such as NeurIPS, ICML, ICLR, ACL, EMNLP, or COLM) and/or work that shaped an LLM development cycle.
  • Experience working closely with researchers and cross-functional stakeholders, including subject-matter experts, to turn ideas into practical training or evaluation assets.
  • Excellent written and oral communication skills and comfort with ambiguity in a fast-paced environment.
  • For Lead level: 3+ years of research experience after the PhD; a record of leading a research agenda to published or deployed results; and experience mentoring or supervising researchers.

Nice To Haves

  • Experience with RL post-training frameworks such as NeMo-RL, verl, OpenRLHF, or TRL, or evaluation frameworks such as Inspect or lm-evaluation-harness.
  • Experience designing LLM-as-a-judge or rubric-based reward and evaluation pipelines and measuring their reliability.
  • Experience with Mixture-of-Experts or long-context models.
  • Experience in law, tax, or other professional, non-verifiable domains, particularly grounding and citation.
  • Experience in experimental design and best practices for working with human data.
  • Familiarity with AI-assisted coding and using coding agents at scale.

Responsibilities

  • Design benchmarks, training datasets, rewards, and training signals to improve the Thomson model family, including for tasks with scarce, delayed, or partly subjective ground truth.
  • Design rigorous experiments and diagnostics, such as ablations, to distinguish algorithmic effects from data, reward, and system artefacts, and carry results into production models.
  • Build and extend the evaluation suite to understand model strengths and weaknesses and inform training objectives and technical trade-offs.
  • Collaborate with Research Engineers on large-scale training and evaluation infrastructure and with subject-matter experts in law, tax, and other professional domains.
  • Coordinate with product teams and other model consumers to understand adoption requirements and drive integration.
  • Publish regularly at top-tier ML and NLP venues and represent Thomson Reuters Labs in the research community.
  • At Lead level, own a research agenda, turn emerging methods into a roadmap for the models, and mentor scientists and engineers.

Benefits

  • Flexible work arrangements, including work from anywhere for up to 8 weeks per year.
  • Flexible vacation and competitive vacation, sick and safe paid time off.
  • Two company-wide Mental Health Days off annually.
  • Access to the Headspace app and Employee Assistance Program.
  • Retirement savings and a competitive 401(k) plan with company match.
  • Tuition reimbursement.
  • Employee incentive programs and possible annual bonus based on enterprise and individual performance.
  • Health, dental, vision, disability, and life insurance programs.
  • Paid holidays, including two company mental health days.
  • Parental leave and sabbatical leave.
  • Flexible Spending and Health Savings Accounts.
  • Fitness reimbursement.
  • Commuter benefits.
  • Adoption and surrogacy assistance.
  • Employee Stock Purchase Plan.
  • Two paid volunteer days off annually.
  • Resources and programs supporting mental, physical, and financial wellbeing.
  • Optional hospital, accident and sickness insurance, paid 100% by the employee.
  • Optional life and AD&D insurance, paid 100% by the employee.
  • Group Legal Identity Theft Protection benefit, paid 100% by the employee.
  • Access to a 529 Plan.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service