Machine Learning PhD Intern: Spring 2027

We are a rapidly growing, Boston-based medical technology company using AI to enable clinicians to make informed decisions on cardiovascular care - leading to better patient outcomes, improved quality of care, and reduced healthcare costs. Come work at Elucid and be part of delivering on our mission to prevent heart attacks and strokes!
This role is hybrid and based out of our Boston office, open to local PhD candidates.
As a leader in cardiac CT imaging, we’re exploring how large language models (LLMs) and vision-language models (VLMs) can improve patient outcomes, make life easier for clinicians and analysts, and even power entirely new products. We’re looking for a motivated PhD-level intern to help us find out. You’ll design, build, and evaluate LLM/VLM agents and workflows, and create proofs-of-concept for real applications of these foundation models. Your work may have the opportunity to join our existing product suite or brand-new standalone product, or lead to publication with our team. If you’re excited to bring foundation model research into clinical care, this is a great opportunity to work at the intersection of innovation and commercial application.
This is a part-time internship (with the option for more full-time hours) starting in January 2027, with at least two in-office days a week on Monday and Wednesday. Our sunny, spacious office is located in the heart of Boston, next to Boston Common.
This is a role for someone who is a:

• PhD student in Machine Learning, Computer Science, Electrical Engineering, or a related field, with a focus on foundation models and an interest in healthcare technology.
• Self-driven researcher who has owned projects from idea to result and enjoys turning ambitious concepts into working proofs-of-concept.
• Clear communicator who works well on a team, partnering with engineers, project managers, and clinical staff.

With these skills and experiences:

• Hands-on experience with Python/PyTorch and tools such as PEFT/LoRA, HuggingFace, or LangGraph (or similar) for model development, fine-tuning, and agentic workflows (AWS experience with EC2, S3, or SageMaker is a plus).
• Experience with image data and complex datasets (multimodal, richly labeled, or requiring substantial preprocessing). Biomedical or volumetric data like DICOMs is a plus, and cardiovascular imaging is best of all.
• Substantial publications in the domain of LLMs or VLMs preferred.
• Experience with human-in-the-loop training, RLHF, or tools like nnInteractive and experience working in an FDA-regulated environment are a plus.

Salary: $60-$65/hour for up to an estimated 20 hours a week.