Sr Scientist Biostatistics

Johnson & Johnson Innovative Medicine

About The Position

The Discovery and Manufacturing Statistics organization at Johnson & Johnson seeks a biostatistician with strong statistical training and applied AI experience. The role combines biostatistics, artificial intelligence, and scientific software development to create trustworthy, validated, and reusable AI-enabled statistical solutions for pharmaceutical research and manufacturing. Solutions should be statistically sound, scientifically interpretable, scalable, and appropriate for decision-making in regulated and validation-sensitive settings. The successful candidate will work with domain experts and technology teams, developing with large language models, retrieval-augmented generation, and agentic workflows. The role supports the integration of scientific intelligence into digital and laboratory environments, emphasizing reproducibility, traceability, and expert oversight. Listed locations include Beerse, Belgium; Leiden, the Netherlands; Prague, Czechia; Maidenhead, United Kingdom; Spring House and Raritan, United States; and Titusville, Pennsylvania. The position also provides scientific and technical leadership by educating and mentoring statisticians and scientists in the effective use of AI, machine learning, and modern computational methods.

Requirements

PhD in Statistics, Biostatistics, Machine Learning, Data Science, Computer Science, Artificial Intelligence, or a related quantitative discipline; alternatively, a Master’s degree with commensurate relevant experience.
Candidates whose primary degree is in Computer Science, AI, or Software Engineering must demonstrate advanced knowledge and practical application of statistical modelling, study design, uncertainty quantification, and model validation.
Ability to communicate technical trade-offs clearly in English and collaborate with colleagues from different scientific and technical backgrounds.
Strong foundations in statistical inference and predictive modelling.
Understanding of study and experimental design.
Experience assessing uncertainty, bias, overfitting, and generalizability.
Experience with model validation, calibration, and performance evaluation.
Proficiency in R and/or Python for statistical analysis.
Ability to explain complex statistical and AI concepts to scientific audiences.
Hands-on experience developing AI or machine-learning solutions.
Experience with generative AI, large language models, retrieval-augmented generation, or agentic workflows.
Demonstrated experience designing end-to-end architectures for production applications, including service boundaries, data flows, API contracts, security, scalability, reliability, and observability, and ability to explain architectural trade-offs.
Strong Python skills and familiarity with building maintainable APIs and services.
Working knowledge of SQL, data pipelines, and data-quality controls.
Experience building modular, tested, and reproducible analytical software using Git.
Understanding of AI evaluation, monitoring, and responsible-use considerations.
Ability to connect AI components beyond the proof-of-concept stage to validated statistical or scientific functions.
Experience moving AI or statistical solutions from prototype to a supported production application.

Nice To Haves

Experience in pharmaceutical, biotechnology, healthcare, or other regulated or validation-sensitive environments.
Experience with manufacturing, discovery, omics, or other complex scientific data.
Experience with Bayesian modelling, design of experiments, mixed models, causal inference, or high-dimensional data.
Scientific modelling, active learning, or optimisation experience, particularly in iterative experimental workflows.
Experience with RAG evaluation, hybrid retrieval, knowledge graphs, and contextualisation of scientific data.
Experience with MCP, agent orchestration, shared model services, and workflows combining automated checks with expert review.
R package development, Shiny, Quarto, interoperable R/Python workflows, FastAPI, or modern web interfaces; familiarity with Posit Workbench or Posit Connect.
Experience with hosted or self-hosted model APIs, LLM serving frameworks such as vLLM, and enterprise deployment environments.
Experience with AI observability, access enforcement, prompt-injection defenses, and responsible handling of sensitive data.

Responsibilities

Make statistical, mechanistic, and predictive models accessible to AI workflows through documented interfaces and shared model services; build Python services and connect approved R and Python tools through APIs and Model Context Protocol (MCP).
Ensure the statistical integrity of AI-enabled solutions, including appropriate treatment of variability, uncertainty, confounding factors, and generalisability.
Design, develop, and deploy AI-enabled scientific applications that integrate statistical methods, machine learning, generative AI, and domain knowledge into scalable, reusable solutions.
Develop agentic workflows that retrieve evidence, invoke scientific models, and support interpretation and proposed next experiments, incorporating controlled tool execution, error handling, and human review.
Prepare, curate, and contextualise scientific data, experimental results, and documents for AI applications; address data quality, metadata, provenance, and uncertainty; and contribute to knowledge-graph and semantic-integration initiatives where relevant.
Collaborate with modelling and laboratory automation teams on iterative workflows connecting experimental results, model updates, and experimental design while maintaining scientific rigour, traceability, and governance.
Work with platform and security teams to deploy and operate applications, including access controls, configuration, logging, monitoring, and validation.
Design end-to-end architectures for production scientific AI applications covering user interfaces, services, APIs, data, models, and deployment; define integration patterns suited to shared enterprise platforms.
Test applications and evaluate retrieval, tool use, and generated outputs; work with domain experts to assess scientific validity, uncertainty, and limitations, and investigate failures.
Contribute reusable statistical and AI capabilities across DMS, including documentation, software components, analytical frameworks, validation approaches, best practices, and technical standards.
Educate and mentor statisticians and scientists in applying AI, machine learning, and modern computational methods.

Benefits

Annual bonus with a target depending on pay grade and location, with the actual amount based on employee and company performance in the previous calendar year, or sales commissions.
Vacation days.
Parental leave for a minimum of 12 weeks.
Bereavement leave.
Caregiver leave.
Volunteer leave.
Well-being reimbursement.
Programs for financial, physical, and mental health.
Service anniversary and recognition awards.
Participation in several insurance plans for employees and, in some locations, eligible dependents, subject to plan terms.