ML Research Engineer / Scientist (Remote)

About the role

Join a research‑driven team building next‑generation AI models that interpret complete CT studies. You will work on foundation models, vision‑language learning, and multi‑finding detection using large‑scale medical imaging data, moving innovations from the lab to clinical practice.

Key responsibilities

• Design, develop, train, and evaluate machine learning models that interpret whole CT studies.
• Research foundation‑model approaches for 3D and volumetric medical imaging at scale.
• Develop vision‑language models linking images with radiology terminology and reporting patterns.
• Build systems that detect and prioritize multiple urgent clinical findings while maintaining safe operating points.
• Design and execute independent experiments, from hypothesis formation to training, evaluation, and analysis.
• Create custom architectures, loss functions, training pipelines, and distributed training solutions using modern deep‑learning frameworks.
• Rigorously analyze model performance and establish reproducible evaluation methodologies for clinically consequential AI.
• Collaborate closely with fellowship‑trained radiologists to translate research findings into practical model improvements.
• Contribute to efforts aimed at regulatory submissions, clinical deployment, and real‑world patient use.
• Partner with ML and software engineering teams to move successful research toward robust, deployable systems.

Required profile

• Strong practical experience with modern machine learning and deep learning, especially using PyTorch for custom architectures and experimentation.
• Deep understanding of machine‑learning architecture design and training dynamics.
• Ability to design and run independent research experiments end‑to‑end.
• Experience working with large‑scale medical imaging data is a strong plus.
• Excellent collaboration skills to work with radiologists and engineering teams.

Required skills

• Python
• PyTorch
• Deep learning frameworks
• Distributed training techniques
• Custom model architecture development
• Vision‑language modeling
• 3D/volumetric learning
• Model evaluation and calibration