Senior Scientist, Data (AI) Scientist, Translational Safety

Position Summary

Senior Scientist – Data (AI) Scientist, Translational Safety to join the Data, Data Science & Artificial Intelligence (DDSAI) – OCMO (Office of Chief Medical Officer) organization helping accelerate drug safety prediction across all stages of drug discovery and development, using advanced AI/ML analytics and multimodal modeling of biological (preclinical and clinical) and RWE data. The Senior Translational AI Scientist will develop predictive models that identify translational-biomarkers and flag compounds with high translational-safety risk, enabling optimal risk minimization supporting patient benefit/risk decisions. One exciting opportunity will be to support the development of AI-enabled reasoning capabilities and Foundation models that connect discovery, preclinical, clinical, and real-world evidence domains to accelerate translational safety predictions. This role will partner closely with pharmaceutical scientists across all phases as well as other Data Scientists from Knowledge Engineering and Data Products to transform harmonized, AI-ready data assets into actionable scientific insights that improve decision-making across the R&D lifecycle.

Mission

Develop scientifically credible AI and machine learning capabilities and models that enable earlier prediction of safety and efficacy outcomes and support closed-loop learning across drug discovery and development.

Strategic rationale (why this role matters)

• Builds the capability for AI-driven, translationally-focused predictive models that identify safety biomarkers and flag high translational-risk compounds earlier in the R&D lifecycle in the forward direction, and also support reverse-translation of AE (Adverse event) Signals from RWE by feeding back causal inference insights to preclinical and clinical.
• Directly supports faster, evidence-based go/no-go and risk-minimization decisions, increasing program productivity and protecting patient safety.
• Enables development of cross-domain Foundation models and AI reasoning that connect discovery → preclinical → clinical → RWE, creating reusable IP and accelerating future projects.

Key responsibilities

• Design, build, validate and deploy AI/ML solutions for translational safety prediction using multimodal data across discovery, preclinical, clinical and real-world evidence (RWE) domains.
• Develop predictive models and AI-reasoning frameworks for translational safety, biomarker identification, mechanistic inference and clinical outcome prediction; evaluate traditional ML, deep learning, causal inference and foundation-model approaches.
• Integrate heterogeneous, high-dimensional datasets (e.g., high-content imaging, phenomics, transcriptomics, proteomics, EHR, claims) to derive novel biological insights that de-risk safety signals and inform portfolio decisions.
• Establish scientific validation: assess biological plausibility, benchmark model performance, produce explainability and validation packages suitable for regulatory and cross-functional review.
• Prototype and advance Foundation-model connectors and AI-enabled reasoning to link discovery → preclinical → clinical → RWE and support closed-loop learning across R&D.
• Collaborate closely with translational scientists, toxicology, clinical safety, PV, Knowledge Engineering, Data Products and external partners to prioritize use cases, operationalize models and drive productization.
• Communicate complex technical methods and results clearly to diverse audiences and stakeholders; maintain reproducible code, documentation and up-to-date versioned repositories.

Qualifications

Education: Ph.D. preferred in Computational Biology, Bioinformatics, Biomedical Informatics, Computer Science, Statistics, Applied Mathematics or a related quantitative discipline; or equivalent experience.

Experience: demonstrated experience applying AI/ML in life-sciences settings (industry or post-doc); typically 2+ years post-PhD or ~3–5 years relevant industry experience.

Domain expertise: track record in translational science, biomarker discovery, safety assessment or related drug-discovery applications.

Publication history or demonstrated contributions to top-tier conferences/journals preferred.

Technical skills: Strong programming proficiency, preferably Python, and experience with AI frameworks (PyTorch or TensorFlow). Deep knowledge of ML/DL methods (Transformers, CNNs, graph networks, self-supervised and multi-instance learning), causal inference and graph analytics. Experience with multimodal representation learning, foundation models, LLMs/GraphRAG and multimodal data fusion. Practical experience analyzing imaging/microscopy, multi-omics and real-world clinical data at scale.

Capabilities & behaviors: excellent analytical thinking, scientific rigor, strong written and oral communication, collaborative cross-functional influence, and the ability to translate domain questions into robust AI solutions.

Nice-to-have: prior experience producing regulatory-acceptable model evidence, operationalizing models into decision workflows, or building reusable model assets for R&D programs.