Postdoctoral Fellow in AI/ML for Maternal and Perioperative Care

Our research integrates physiological signals, ultrasound imaging, clinical text, and electronic health records to develop and rigorously evaluate methods for clinical prediction. Strong computational researchers from a range of quantitative disciplines are welcome.

The fellow will develop original machine learning methods using large-scale clinical datasets, with applications to maternal and perioperative health. Potential research directions include physiological time-series modeling, computer vision and representation learning for ultrasound, clinical text analysis, and multimodal integration of signals, images, text, and structured electronic health records. The position emphasizes both methodological innovation and rigorous evaluation. The fellow will formulate research questions, design experiments, implement and validate models, investigate limitations and generalizability, and communicate findings through publications and presentations. Specific projects will be shaped jointly around the fellow’s strengths and interests, aligned with the lab’s funded research priorities. The fellow will work closely with clinical and computational researchers and have opportunities to lead projects, contribute to collaborative research, and mentor junior team members. Prior clinical experience is not required, and candidates are not expected to have expertise across every modality.

This is a full-time, Boston-based position with a hybrid working arrangement that includes on-site participation. The initial appointment is for one year, renewable annually, within a research program funded for five years. Salary follows NIH postdoctoral guidelines and is based on experience and institutional policy. International applicants are welcome; J-1 visa sponsorship is available for eligible candidates. Applications will be reviewed on a rolling basis until the position is filled.

• A PhD in computer science, engineering, statistics, applied mathematics, physics, or a related quantitative discipline, completed before the start date.
• Strong Python programming skills and hands-on experience developing and evaluating machine learning or deep learning models.
• Research experience in at least one relevant area, such as time-series analysis, signal processing, computer vision, representation learning, natural language processing, or multimodal learning.
• Evidence of original scientific contributions, thoughtful experimental design, rigorous model validation, and the ability to critically assess and verify one’s own code and analyses.
• Strong scientific communication skills and the ability to work independently and collaboratively are important.
• Experience with medical imaging, physiological signals, or clinical datasets is welcome but not required.
• We value strong computational foundations, scientific curiosity, and the capacity to learn new methods and application domains.