Clinical Data Scientist

Xenon7•Ohio City, OH

1d•Hybrid

About The Position

The Clinical Data Scientist will lead the development and application of digital twins to improve clinical trial design and support evidence-based clinical development decisions. The role prioritizes digital-twin, simulation, and AI/ML expertise, supported by clinical-development experience and close partnership with trial-design experts. The position works within a multidisciplinary clinical development environment, partnering with clinical scientists, physicians, biostatisticians, pharmacometricians, biomarker experts, data scientists, data engineers, and other stakeholders. The role translates clinical development questions into quantitative strategies and delivers interpretable analyses to inform protocol design, patient selection, endpoint strategy, dosing, operational planning, and decisions under uncertainty. The primary capability is developing digital twins to simulate patients, disease trajectories, treatment response, and clinical trials. The successful candidate brings strong digital-twin, simulation, and AI/ML expertise and sufficient clinical-development knowledge to collaborate effectively with trial teams. Deeper trial-design expertise may be developed through collaboration with clinical scientists, biostatisticians, and pharmacometricians. This is a full-time contractor role for an enterprise project engagement, outsourced via Xenon7. The position is based in the Indianapolis, Indiana metro area and is hybrid with three onsite days; regional or Eastern Time candidates who can travel onsite are also considered.

Requirements

PhD in data science, statistics, biostatistics, pharmacometrics, computer science, biomedical informatics, engineering, mathematics, operations research, or a related quantitative field.
Demonstrated experience developing digital twins, virtual patients, disease-progression models, clinical trial simulations, synthetic controls, or closely related computational models.
Strong experience applying machine learning, statistical modeling, causal inference, or simulation to biomedical, healthcare, pharmaceutical, or other complex scientific data.
Strong programming skills in Python, R, or a comparable analytical language, with experience using version-controlled and reproducible workflows.
Sufficient clinical-development knowledge to frame relevant questions and collaborate effectively with clinical and quantitative trial experts.
Ability and interest to deepen clinical trial-design knowledge through partnership with clinical scientists, biostatisticians, and pharmacometricians.
Strong written and verbal communication skills and demonstrated effectiveness in cross-functional scientific collaboration.

Nice To Haves

Experience supporting clinical trial design, protocol development, clinical development strategy, model-informed drug development, or regulatory-facing quantitative evidence.
Familiarity with trial-design topics such as endpoint strategy, sample size or power assumptions, patient selection or enrichment, longitudinal outcomes, missing data, adaptive designs, or decision criteria.
Knowledge of digital-twin architecture and validation, supported by expertise in causal inference, Bayesian modeling, survival analysis, longitudinal modeling, clinical trial simulation, or synthetic controls.
Experience with clinical data standards and common clinical-development data sources.
Experience with biomarker, imaging, omics, digital health, or real-world data and integration of multiple data modalities.
Experience with scalable cloud-based analytical environments.
Track record of translating analytical work into decisions, reusable capabilities, publications, presentations, or other measurable scientific impact.

Responsibilities

Lead the design, development, and evaluation of digital twins of patients, disease trajectories, treatment responses, or clinical trials using clinical trial and relevant external data.
Combine mechanistic, statistical, causal, and machine-learning approaches to create fit-for-purpose digital twins that can be simulated under alternative populations, interventions, dosing strategies, endpoints, or trial designs.
Define use cases and contexts of use with clinical and quantitative experts so each digital twin addresses a specific trial-design or development decision.
Assess calibration, validation, robustness, transportability, uncertainty, and practical relevance using appropriate internal and external evaluation strategies.
Develop reproducible workflows and reusable digital-twin components for application across programs or therapeutic areas.
Partner with trial teams to frame development questions and evaluate design options, including target populations, eligibility criteria, enrichment strategies, comparators, endpoints, visit schedules, dosing strategies, and operational tradeoffs.
Use digital twins and complementary simulation methods to compare plausible scenarios, characterize uncertainty, and support model-informed decision-making.
Work with statisticians and pharmacometricians to complement established trial-design methods and align analyses with clinical and regulatory objectives.
Communicate assumptions, limitations, and decision-relevant insights clearly to technical and non-technical stakeholders.
Analyze and integrate clinical trial data and, where relevant, biomarkers, imaging, digital health data, omics, patient-reported outcomes, and real-world data.
Partner with data engineering and platform teams to create analysis-ready datasets, data-quality checks, data lineage, and reproducible pipelines.
Contribute to scientific documentation, presentations, publications, and responsible adoption of digital-twin methods across clinical development.
Define the intended use, decision context, success criteria, and limitations of each model before development.
Ensure models are interpretable, auditable, reproducible, and supported by evidence appropriate to their intended use.
Apply appropriate governance, privacy, security, and scientific-review requirements for clinical data and analytical outputs.
Distinguish exploratory analyses from validated decision-support evidence and avoid overclaiming model capability.