Bristol Myers Squibb logo
2 hours ago
Remote friendly (Princeton, NJ)
United States
Other
Position Summary
- Join the Drug Development Data Science and Advanced Analytics (DSAA) team to drive and shape the global drug development process.
- Serve as a recognized scientific and technical leader within DSAA, defining and championing exploratory data science approaches and platforms across early-to-late phase drug development.
- Collaborate closely with Biostatistics, Translational and Clinical Scientists, and other cross-functional partners to integrate data science work into development, regulatory, and commercial strategy.

What You’ll Do
Digital Health & Wearable Data Science
- Lead strategy and hands-on execution for wearable/sensor-derived time-series data, including Python pipelines for QC, preprocessing, sensor artifact removal, imputation, and feature engineering.
- Develop and validate longitudinal sensor models (e.g., frequency/time-frequency representations, digital filtering, representation learning, Transformers/ensembles) with rigorous explainability.
- Implement statistically rigorous methods for repeated-measures/longitudinal data (e.g., mixed-effects/hierarchical models; missingness and within-subject dynamics).
- Quantitatively characterize physiological/clinical measures (e.g., accelerometry/actigraphy, HRV, SpOâ‚‚) for disease progression, patient subtyping, or treatment response with regulatory-grade evidence aims.
- Oversee and evaluate third-party analytics providers and vendor-derived digital biomarker outputs.
- Champion evaluation standards, reproducible research, and scalable engineering principles.

Broader Multi-Modal Data Science
- Drive biomarker analyses across multimodal clinical, digital health, and omics datasets (genomics, proteomics, imaging, flow cytometry, other high-dimensional biomarker data).
- Lead novel computational methods for patient segmentation, predictive biomarker discovery, and precision medicine.
- Provide senior input into statistical analysis plans (SAPs) for exploratory biomarker and digital health analyses.
- Perform/oversee rigorous analyses of high-dimensional data (e.g., gene expression, sequencing, imaging features) with interpretability.
- Integrate, mine, and visualize diverse high-dimensional datasets across therapeutic areas/phases; develop new analytical frameworks when needed.
- Formulate, implement, test, and validate predictive models and scalable automated processes to deliver results across programs.
- Apply modern ML/AI methods (AI/ML, deep learning, NLP, causal ML, explainable AI) to accelerate drug development.
- Influence drug development strategy, including predictive biomarkers, novel trial designs, and precision medicine.

Leadership, Strategy & Cross-Functional Influence
- Shape DSAA vision/roadmap for digital health and multi-modal data science.
- Manage/develop a small team of data scientists (if applicable).
- Build partnerships with Biostatistics, Translational Medicine, Clinical Development, Regulatory Affairs, and IT/Engineering.
- Lead cross-functional discussions and represent DSAA in project team meetings to influence decisions.
- Establish best practices and mentor junior/mid-level data scientists.
- Communicate complex analyses to technical and non-technical audiences.
- Manage competing priorities to deliver outputs on timeline.

Key Requirements
- Ph.D. in a relevant quantitative field (e.g., Computational Biology, Biostatistics, Statistics, Biomedical Engineering, Computer Science) + 8+ years academic/industry experience; or Master’s + 10+ years industry experience.
- Deep, hands-on digital health data science expertise with wearable/sensor time-series data (QC, preprocessing, artifact handling, imputation, feature engineering for accelerometry/actigraphy, HRV, SpOâ‚‚) with validated, production-quality outputs.
- Demonstrated mastery analyzing clinical trial or EHR-generated data for pharma R&D.
- Track record driving statistical and AI/ML innovation across multiple data modalities.
- Strong Python; experience leading production-quality, modular/scalable pipelines; Git/version control; collaborative software development.
- Expertise applying statistical/ML models to high-dimensional time-to-event, longitudinal, and multivariate outcomes.
- Familiarity with clinical trial design, drug development processes, and biomarkers’ role in regulatory/clinical decisions.
- Proven ability to influence strategy via rigorous data-driven analysis.
- Ability to lead/mentor/collaborate with multidisciplinary teams and manage multiple concurrent priorities.
- Excellent communication, presentation, visualization, and ability to convey complex concepts to diverse audiences.
- Capable of establishing high-trust relationships across the organization.

Preferred Qualifications
- Experience with genomics, proteomics, imaging, flow cytometry, or immunobiology datasets (highly preferred).
- Experience with NLP (highly preferred).
- Survival analysis/time-to-event modeling (highly preferred).
- Causal ML and explainable AI (highly preferred).
- Knowledge of molecular biology and disease pathways.
- Experience/perspective on novel trial designs (adaptive, platform, biomarker-enriched).
- Sleep analytics, circadian cosinor modeling, or movement physics (quaternions/Euler angles/orientation estimation).
- Experience overseeing/integrating third-party analytics partnerships and evaluating vendor outputs.
- Scalable compute/deployment experience (e.g., AWS, multi-GPU, parallelization).
- People management or formal scientific leadership experience (plus).

Compensation Overview
- Princeton, NJ, US: $218,740 - $265,060 (base compensation range). Additional incentive cash and stock may be available (eligibility-based).
- Benefits include health coverage, wellbeing support programs, financial wellbeing/protection (e.g., 401(k), disability, life insurance), and paid time off (details vary by employee category).

Application Instructions
- If applying, review compensation/benefits eligibility as listed for the role and location; apply anyway if the role interests you even if you don’t perfectly match every requirement.