Machine Learning Engineer

Cardiovascular disease is the #1 cause of morbidity and mortality in the world. Much of this could be prevented with better access to specialist care. Take stroke as an example: any delay in treatment can lead to permanent disability or death. However, due to a lack of specialist surgeons, the most effective intervention can only be performed in 2% of US hospitals. For patients who present to one of the 98% of hospitals that do not offer the surgery, treatment is either significantly delayed or not offered at all because timely transfer is not feasible.

Our mission is to bring state-of-the-art vascular intervention to anyone, anytime, regardless of their location. Our team of medical clinicians, roboticists, and machine learning experts are working to bridge this gap by building the world’s first remotely-operated, semi-autonomous endovascular surgical robot.

We’ve already done what nobody else could—using our system, doctors from around the world were able to remotely perform this procedure from as far as 8000 miles away. We have now successfully performed first-in-human cases, including a remotely operated procedure, demonstrating the potential of our technology to revolutionize access to life-saving interventions. We now need your help to bring this technology out of the laboratory and into hospitals everywhere.

We’re looking for someone to continue leveraging our vast trove of medical imaging data in order to train and deploy deep neural network models. These models enable our surgical robot to understand and reason about both our robot and the patient’s anatomy, which ultimately gives doctors the insight and control necessary to quickly and safely complete the procedure.

• At least 2 years of machine learning engineering experience (level will be commensurate with your experience)

• Experience developing high-quality software, ranging from design and implementation to testing and deployment

• Expertise with Python

• Experience training image-based deep neural networks, including

  • Deep neural network libraries such as PyTorch

  • Defining training and validation datasets

  • Using data augmentations during training

  • Selecting loss functions and metrics

  • Cloud-based data and training

  • Conducting large-scale experiments to determine actionable improvements

• Eagerness to learn on the job, iterate fast, and collaborate

• Experience developing and deploying neural networks for physical systems, such as robots and autonomous vehicles

• Experience with medical imaging data such as x-rays, CTs, and MRIs

• Experience bridging the sim-to-real gap