Senior Data Scientist (NLP & Applied AI)

Role overview Develop NLP and applied AI systems that make large collections of scientific articles easier to search, interpret, and use in downstream applications. You’ll own content modeling from evaluation through delivery, working hands-on in both exploratory analysis and production code, and collaborating with editorial, product, and engineering partners.

Responsibilities - Build enrichment pipelines that extract entities, classifications, claims, and summaries from scientific full text. - Evaluate classical NLP methods alongside transformers, embeddings, retrieval, and LLM-based approaches to choose suitable methods for each task. - Create evaluation sets and metrics, then use results to improve quality, manage cost, and detect performance changes. - Contribute model-grounding and evaluation expertise to tool-using AI applications built on enriched content. - Bring modeling evidence into product decisions and turn feedback from editors, product managers, and engineers into specific work. - Produce clear, maintainable Python code that can be extended by engineers and other data scientists.

Requirements - Strong background in modern and classical NLP, including LLMs, transformers, embeddings, retrieval, named-entity recognition, classification, or sequence labeling. - Experience building evaluations and using their results to inform modeling decisions. - Clean, maintainable Python skills spanning exploratory notebooks and production repositories. - Ability to compare modeling options and explain choices using evaluation results, cost estimates, and operational considerations. - Sound judgment about how to respond when model performance drifts.

Nice to have - Experience with scholarly or scientific text, production LLM workloads with cost and latency limits, or agentic applications involving tool use, multi-step reasoning, safeguards, and trajectory evaluation. - Familiarity with AWS services such as S3, Batch, Lambda, and SageMaker, or Parquet and Iceberg data lake patterns.

Benefits and work setup - The source describes a comprehensive benefits package and meeting-free Friday afternoons for focused work and development. Compensation varies by location and candidate qualifications.