About this role
Based in Boston, United States, Harvard University is hiring for a Postdoctoral Fellow – Large Language Models position, working on-site. The role sits in Artificial Intelligence & Data, which shapes the kind of work involved. This is an in-office job in Boston, not a remote one. Pitched at entry-level, the role expects that amount of experience behind you. If Boston works for you, the full description below has the rest.
Your Career Hunter matched this listing against the employer's careers page on Nov 12, 2026.
Job Overview
Job description provided by Harvard University.
The Kohane Lab at Harvard Medical School's Department of Biomedical Informatics seeks a highly motivated Postdoctoral Fellow to work on large language models.
The fellow will contribute to research on delivering primary care using large language models, a cutting-edge area of medical AI.
Work involves research and development using large language models for medical text processing.
The fellow will collaborate with a multidisciplinary team to design and implement novel algorithms and models for medical text analysis.
The role includes analysing large-scale medical datasets, extracting relevant information, and improving the accuracy and efficiency of medical text processing.
Developed algorithms and models will be evaluated and validated through rigorous experimentation and benchmarking.
Findings will be published in high-impact scientific journals and presented at conferences and seminars.
The fellow will also help mentor and supervise other researchers in the lab on projects related to medical text processing.
Applicants need a PhD in biomedical informatics, computer science, computational linguistics, or a related field, with a strong publication record.
Strong hands-on expertise with large language models for medical text, plus proficient programming in Python and R, is required. Applications go by email to jobs@zaklab.org, including a cover letter, detailed CV, list of references, and copies of the most significant publications or preprints.