[Remote] Senior Applied Scientist, Document Understanding

Note: The job is a remote job and is open to candidates in USA. Thomson Reuters is a global leader in providing trusted content and technology to professionals across various sectors. They are seeking a Senior Applied Scientist focused on designing and deploying document understanding systems that enhance their legal products. The role involves working on semantic chunking, document enrichment, and knowledge graph construction to deliver foundational intelligence for multiple product teams.ResponsibilitiesDesign and deploy semantic chunking models for lengthy, non-uniformly structured legal documents with adjustable granularity across use casesBuild document enrichment systems that classify documents according to legal and customer-defined taxonomies and extract rich metadataDevelop LLM-based knowledge graph construction pipelines that extract and link citations, entities, and legal concepts across diverse legal contentBuild scalable synthetic data generation systems for model training, multi-hop query simulation, and hallucination-free answer generationApply knowledge distillation techniques to compress large models into latency-constrained, production-ready SLMsDesign evaluation frameworks — component-level and end-to-end — using expert annotation and synthetic dataDrive independent technical decisions on chunking strategy, classification approach, knowledge extraction methods, and multi-document reasoning architecturePartner with engineering on delivery, reliability, and scale across multiple product linesContribute to published research at venues such as ACL, EMNLP, ICLR, NeurIPS, SIGIR, and KDD, and to intellectual propertySkillsPhD or Master's in Computer Science, AI, NLP, or a related field5+ years of post-degree industry experience shipping document understanding, information extraction, or knowledge graph systems into production — not research-only experiencePublications at ACL, EMNLP, ICLR, NeurIPS, SIGIR, KDD, or equivalentExperience leading through influence in an applied research settingProduction Python and experience with PyTorch, Hugging Face Transformers, and DeepSpeedDocument layout analysis and semantic chunking beyond fixed-size or paragraph-based methodsHierarchical, multi-label document classification with domain-specific and customer-defined schemasEntity recognition and linking, relation extraction, citation parsing, and knowledge graph construction from unstructured textLLM-based information extraction, few-shot and multi-task learning, and post-trainingKnowledge distillation, model compression, and SLM deployment under latency constraintsSynthetic data generation for NLP: query-answer generation with verification and scalable data augmentationAnnotation workflow design and evaluation framework development for document understanding tasksLegal document understanding, legal information extraction, or legal AI applicationsComplex document structures common in legal content: nested hierarchies, cross-references, non-uniform formatting, and embedded elementsRetrieval, QA, or analysis systems over large document collectionsKnowledge graph frameworks for legal or enterprise applicationsRAG and agentic workflows for enterprise knowledge systemsAzureML or AWS SageMakerBenefitsFlexibility & Work-Life BalanceCareer Development and GrowthIndustry Competitive BenefitsCultureSocial ImpactMaking a Real-World ImpactMarket competitive health, dental, vision, disability, and life insurance programsCompetitive 401k plan with company matchCompetitive vacation, sick and safe paid time offPaid holidays (including two company mental health days off)Parental leaveSabbatical leaveOptional hospital, accident and sickness insurance paid 100% by the employeeOptional life and AD&D insurance paid 100% by the employeeFlexible Spending and Health Savings AccountsFitness reimbursementAccess to Employee Assistance ProgramGroup Legal Identity Theft Protection benefit paid 100% by employeeAccess to 529 PlanCommuter benefitsAdoption & Surrogacy AssistanceTuition ReimbursementAccess to Employee Stock Purchase PlanCompany OverviewThomson Reuters (TSX/NDAQ: TRI) informs the way forward by bringing together the trusted content and technology that people and organizations need to make the right decisions. It was founded in 2008, and is headquartered in Toronto, Ontario, CAN, with a workforce of 10001+ employees. Its website is https://www.tr.com.Company H1B SponsorshipThomson Reuters has a track record of offering H1B sponsorships, with 1 in 2026, 13 in 2025, 12 in 2024, 5 in 2023. Please note that this does not guarantee sponsorship for this specific role.