RWD Data Scientist

Real-world data projects are different from trial analyses: large, messy, observational datasets, where the study design and the handling of confounding decide whether the result means anything. We support researchers and organizations running cohort, case-control and treatment-effect studies on claims, registry and EHR data.

The RWD Data Scientist leads those projects: defines the study population and variables, builds the data pipeline, applies the right causal and modeling methods, and reports to STROBE, RECORD or the sponsor's standard.

You have worked with real-world datasets at scale and know their failure modes: immortal time bias, confounding by indication, missingness that is not random. You are as careful about the data pipeline as about the model, and you write code that someone else can run.

You want substantial, well-defined projects with the scope and fee agreed before you start, alongside your own research or practice.

• Design observational studies on RWD: population definitions, exposure and outcome algorithms, follow-up and censoring.
• Build reproducible data pipelines in R or Python from raw extracts to analysis-ready datasets, documented.
• Apply causal inference methods: propensity scores, IPTW, target trial emulation, instrumental variables where appropriate.
• Fit and validate predictive and survival models on large datasets.
• Report to STROBE, RECORD or sponsor standards, with limitations stated plainly.
• Advise on feasibility and data requirements at scoping.

• PhD in Epidemiology, Biostatistics, Health Informatics or Data Science.
• Six or more years of applied work with claims, registry or EHR data.
• Expert in R or Python for data engineering and modeling; SQL for extraction; familiarity with OMOP or similar common data models.
• Strong grounding in causal inference and observational study design.
• Experience reporting RWE studies for publication or regulatory and payer audiences.
• Clear written English.
• Based in the United States.

• Machine learning for prediction with proper validation.
• Experience with FDA RWE guidance or ISPE good practice.
• Publications using real-world data.

This is a partnership, not a salaried position. Our core team of PhD-level statisticians is in-house; for specialized projects we bring in a partner statistician, also PhD-qualified and based in the United States, who works under our management, our confidentiality terms and our review. When a project that matches your expertise comes in, we bring you into it. The scope and your fee for that project are agreed before any work begins, so there is never an open-ended commitment on either side. A Research Coordinator handles client contact, scoping and timelines, so your time goes to the statistics. Every project is confidential, and so is your application.