Director, AI Research

AstraZeneca•Boston, MA

8d•Hybrid

About The Position

The Director, AI Research will define and deliver an AI research agenda that improves clinical-development decisions and outcomes for patients across Oncology and BioPharmaceuticals. The role works at the intersection of advanced AI, clinical-trial methodology, and biomedical science, translating methodological innovation into capabilities that inform clinical-development strategy. Research may span foundation models, multimodal learning, agentic systems, post-training methods, and evaluation science, addressing topics such as treatment-effect heterogeneity, predictive biomarker discovery and validation, patient stratification, time-to-event outcomes, cross-trial generalization, and responsible AI evaluation and governance. This director-level position combines original research and method development with reproducible engineering, scientific leadership, external dissemination, team mentorship, and responsible-AI governance. The Director will partner across the organization and with external collaborators, helping ensure advanced AI systems are reliable, reproducible, scientifically sound, and aligned with patient safety and regulatory expectations.

Requirements

PhD in Computer Science, Machine Learning, Biostatistics, Biomedical Informatics, Computational Biology, or a closely related quantitative field, or an MSc with an equivalent record of research and industry impact.
Eight years of experience in original research and technical impact, demonstrated through influential publications, clinical evidence, deployed systems, patents, widely used software or datasets, or comparable contributions.
Experience leading substantial research programs end to end and making scientific or technical decisions that shaped a broader portfolio, platform, or organizational direction.
Ability to set direction under uncertainty, challenge assumptions, and influence senior partners.
Expertise in relevant research topics, such as unsupervised learning, dimensionality reduction, causal inference, survival or longitudinal analysis, Bayesian modeling, multimodal learning, domain adaptation, or model interpretability.
Comfort working with AI agent harnesses, learning new approaches quickly, and investigating what makes agents useful, reliable, and safe.
Strong Python and scientific-computing skills, with experience using a modern ML framework such as PyTorch, JAX, or TensorFlow.
Ability to explain technical findings, tradeoffs, and limitations to clinical, statistical, regulatory, engineering, and leadership audiences.

Nice To Haves

Experience with biomedical data such as clinical text, EHR or trial data, biomedical imaging, genomics, transcriptomics, spatial biology, or real-world data.
Hands-on experience with large language models or agentic systems, including fine-tuning or post-training, retrieval-augmented generation, tool use, agent harnesses, multi-step orchestration, evaluation, or inference and serving.
Familiarity with clinical data standards, regulated validation, GxP, software as a medical device, model-risk management, or interactions with health authorities.
Contributions to scientific consortia, reporting or evaluation standards, open-source communities, or other work that has influenced practice beyond a single project.
Experience mentoring scientists or engineers and providing technical leadership in a global, matrixed organization.

Responsibilities

Set and execute a research strategy focused on high-impact clinical-development challenges and measurable benefits for patients and the pipeline.
Invent and refine AI and statistical-learning methods for randomized clinical trials and related evidence, including methods for heterogeneous treatment effects, predictive biomarkers, time-to-event outcomes, patient stratification, and cross-trial generalization.
Design, build, and rigorously evaluate agentic AI systems for clinical-development use cases, incorporating retrieval, tool use, code execution, memory, multi-step workflows, and human oversight.
Write high-quality Python and build reproducible research workflows using modern ML frameworks, version control, testing practices, and cloud or high-performance computing environments.
Serve as a senior scientific and technical partner, shaping clinical strategy, evaluation approaches, platform capabilities, and investment decisions across the organization.
Share research through peer-reviewed publications, reusable tools, patents, scientific collaborations, and conference presentations; represent the company with partners, consortia, regulators, and the broader research community.
Cultivate a rigorous research culture through structured scientific review, constructive challenge, and mentorship of scientists and engineers.
Establish robust evaluation, governance, and responsible-use frameworks for advanced AI systems in clinical development, aligned with patient safety and regulatory expectations.