Senior Data Scientist – LLM, Clinical NLP

• Build and improve LLM- and transformer-based pipelines extracting structured data from clinical notes
• Develop named entity recognition, text classification, and relation extraction solutions
• Evaluate models using gold-standard datasets, error analysis, and trusted metrics
• Fine-tune, deploy, and monitor models in production
• Work with Medical and Product teams to translate clinical questions into data and model requirements
• Present results and trade-offs clearly to technical and non-technical audiences
• Apply language models to unstructured clinical text for research-grade healthcare data products

• Master's degree or PhD in a quantitative field such as computer science, machine learning, data science, biostatistics, or similar
• Around 5 years of hands-on data science or ML work, including models taken to production on messy real-world data
• Hands-on NLP with transformer encoders, including fine-tuning BERT / RoBERTa-family models and biomedical or clinical variants such as BioBERT or ClinicalBERT
• Experience using Hugging Face Transformers for named entity recognition, classification, or relation extraction
• Hands-on work with generative LLMs, including prompt design and structured outputs, LoRA / PEFT fine-tuning, RAG, and evaluating LLM outputs at scale
• Rigorous model evaluation skills, including gold sets, error analysis, and LLM-as-judge
• Strong Python and SQL
• AI-native use of tools such as Claude and Cursor
• Fluent English and confidence presenting to cross-functional and non-technical stakeholders
• Experience with EHR data or clinical notes and familiarity with ICD, CPT, or RxNorm strongly preferred
• PySpark and Databricks, including MLflow, nice to have
• Experience training, serving, and monitoring models on AWS, especially Amazon SageMaker, nice to have
• R or TypeScript, nice to have
• Ophthalmology, urology, or oncology data experience, nice to have
• Medical imaging data experience, nice to have
• Real-world evidence studies or peer-reviewed publications, nice to have

• Fully paid licenses for Cursor, Claude Pro, etc.
• Remote-first work arrangement with total autonomy
• No filler meetings, Jira bloat, or micromanagement
• Direct impact and face-to-face work with the CEO, CTO, VPs, and client leadership
• Continuous growth and strong knowledge-sharing culture
• Work alongside elite engineers shipping systems that drive real business decisions