ML Research Engineer / Scientist (Boston, hybrid)

a2z Radiology AI•Boston, MA

1d•Hybrid

About The Position

We are building a foundation model that reads entire CT studies, similar to how a radiologist does, drafting the findings into a report. This role focuses on improving the speed and accuracy of this model. The research directly impacts patient care, as the models undergo FDA clearance and are deployed in hospital reading rooms, with experiments potentially influencing clinical practice within months. The position involves working with large-scale, real-world CT datasets paired with radiology reports, and training models from day one. The team consists of builders and researchers who have a track record of shipping FDA-cleared products and collaborating with radiologists across various specialties.

Requirements

Proficiency in PyTorch, including custom architectures, training loops, and distributed training.
Deep understanding of why loss functions and architectures work, not just how to implement them.
Ability to design and execute experiments independently.
Commitment to rigor and reproducibility in model development.
Authorization to work in the United States.

Nice To Haves

Experience with medical imaging, 3D, or volumetric data.
Experience with vision-language models or self-supervised learning.
Background with DICOM, CT, or radiology.
A PhD or publications.

Responsibilities

Build and improve a foundation model that reads entire CT studies.
Develop models for vision-language learning that connect visual data to radiological language.
Create a single model capable of identifying multiple urgent findings simultaneously and operating safely.
Conduct training and evaluation at a large scale.
Own models and experiments end-to-end, from conception to deployment.
Collaborate with ML and software engineers and fellowship-trained radiologists.