Job description:
Associate Applied ML Specialist
POSITION OVERVIEW
Federated Learning and Auto Segmentation of Prostate Cancer Using the FL4Health Platform
As an Associate Applied ML Specialist, you will develop reusable software to apply and scale research breakthroughs in machine learning and AI. The position is a one-year contract with a secondment to the Haider Lab at the Lunenfeld-Tanenbaum Research Institute (LTRI), affiliated with the University of Toronto and Mount Sinai Hospital (SHS) on 600 University Avenue.
You will collaborate with Vector’s AI engineering team and data scientists, a clinician scientist in medical imaging, research fellows, and other doctoral and postdoctoral fellows at the Sinai Health System site at 600 University Avenue.
Over the year long engagement, you will focus on the refinement of an existing model that detects and segments prostate cancer on MRI using nnUnetv2 and MONAI as well as further develop a platform where this model can be extended to use Vector Institute’s FL4Health preform.
RESPONSIBILITIES
- Deliver software and models for AI development projects following an agile methodology with teams that can include Vector Applied ML specialists, Applied ML scientists, professional staff, researchers, and industry sponsors.
- Design and develop reference applications and model templates that apply AI research to real-world problems using public and private data sets, providing software for training, testing, operating, and updating models.
- Develop and write the software framework for testing the prostate cancer detection model with FL4Health framework and test it on a series of multi-institutional data sets.
- Design, develop, and contribute to open-source software that applies AI research techniques to common problems.
- Develop prototypes and proof-of-concepts.
- Report and present software development outcomes for team efforts, including demos, status, and results.
SUCCESS MEASURES
- Timely development and delivery of high-quality software and AI models in support of both internal and external development stakeholders.
- Effective AI knowledge-transfer to internal and external stakeholders through clear communication of software and AI model development progress, demos, and results presentations.
- Active and constructive participation in diverse and interdisciplinary teams, contributing to a positive team environment.
PROFILE OF THE IDEAL CANDIDATE
- Bachelor's degree in computer science, mathematics, electrical engineering, or a related discipline.
- Experience applying machine learning research to novel problems and data sets.
- Broader health sector, medical imaging, and radiomics experience is preferred.
- Demonstrated experience in machine learning and statistics.
- Demonstrated data engineering background working with messy real-world data.
- Demonstrated analytical skills to identify signals in data and feasibility of prediction.
- Knowledge and understanding of Python and preferably C/C++ and C#.
- Software engineering background with experience in production deployment is preferred.
- Experience working on teams that land high-quality software successfully in production is preferred.
- Experience with parameter and architecture tuning of deep learning algorithms is preferred.
- Experience in leading open-source deep learning software frameworks (PyTorch, Tensorflow, JAX, and CUDA) is preferred.
- User interaction/design expertise with Human-Computer Interaction (HCI) engineering considered an asset.
At the Vector Institute, we are committed to driving excellence and leadership in Canada’s knowledge, creation, and use of AI to foster economic growth and improve the lives of Canadians. We strive for greater inclusion in the programs and culture that we build by welcoming and encouraging applications from all qualified candidates. This includes but is not limited to applicants who are indigenous, 2SLGBTQIA+, racialized persons/visible minorities, women, and people with disabilities.
If you require an accommodation at any point throughout the recruitment and selection process, please contact hr@vectorinstitute.ai and we will happily work with you to meet your needs.
