Different routes, clearly separated.

Do research with the lab full time.

We look for researchers who can take responsibility for a difficult question, build the system needed to study it, and communicate the result precisely. The lab is based in Harvard Medical School's Department of Biomedical Informatics and works across clinical settings and real care workflows.

• Postdoctoral fellow, Reason

  2–3 years

  Lead research toward AI that predicts each patient's future: models that learn from a patient's scans, notes, labs, and history over time, with every prediction tied to its evidence.

  REQUIREMENTS · PhD (completed or expected within a year) in computer science, machine learning, biomedical informatics, or a related field; first-author work on clinical prediction, longitudinal or multimodal patient data, or representation learning; and strong implementation skills.

  Apply →

• Postdoctoral fellow, Test

  2–3 years

  Lead research toward a simulator of an entire hospital: clinical agents, simulated patients and hospital environments, and evaluations that show whether AI is ready before it meets a patient.

  REQUIREMENTS · PhD (completed or expected within a year) in computer science, machine learning, biomedical informatics, or a related field; first-author work on agents, simulation, evaluation, or human–AI collaboration; and strong implementation skills.

  Apply →

• Postdoctoral fellow, Assist

  2–3 years

  Lead research toward a portable surgical assistant: robot learning for surgical and bedside procedures, learning from demonstration and procedural video, with surgical collaborators and real hardware.

  REQUIREMENTS · PhD (completed or expected within a year) in robotics, computer science, or a related field; first-author work in robot learning, manipulation, or control with real-robot experiments; and strong implementation skills. Prior medical experience is not required.

  Apply →

• Predoctoral researcher

  1–2 years

  Work full time with the lab before doctoral training, building model-training pipelines, evaluation systems, experiments, and publications within Reason, Test, or Assist.

  REQUIREMENTS · Bachelor’s or master’s degree in a quantitative field; substantial programming experience; and evidence of research potential. Applicants must be based in the United States.

  Apply →

Apply through Harvard's AI in Medicine PhD track.

The lab takes PhD students through the Artificial Intelligence in Medicine (AIM) PhD track in the Department of Biomedical Informatics at Harvard Medical School. Admissions run entirely through the program: apply there directly, and you can list Pranav as your first preference among faculty you want to work with. Please don't email Pranav about PhD admission; applications are read through the program, and email doesn't help.

If you want research experience before a PhD, the two paths into the lab are the predoctoral researcher role above and, for eligible students and physicians, the Medical AI Bootcamp below. Both are competitive: predoc applications are reviewed monthly on a rolling basis, and both use the structured forms on this page, not email.

Already admitted to a PhD program at Harvard or MIT? Reach out right away: rotations with the lab are welcome, and so are conversations about doing your PhD here.

AIM application and FAQs ↗

A distinct mentored research program.

The Medical AI Bootcamp has its own eligibility rules, application, selection process, and reviewer access. It is not an application for a postdoctoral, predoctoral, or PhD position.

View the Medical AI Bootcamp →