Principal Machine Learning Engineer Reconstruction Quantitative Imaging
MidjourneyVisit Midjourney website
Community-funded AI research lab and company providing image and video generation tools.
South San Francisco, United States
About Midjourney
Midjourney, Inc. provides image-generation services through its website and Discord-based services, enabling users to generate and edit images and create videos.
Skills
Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
You will build machine-learning components that improve reconstruction quality, speed, robustness, and quantitative accuracy. You will define training and evaluation pipelines, datasets, and metrics; productionize models with monitoring and safe fallbacks; develop hybrid physics-informed algorithms; and create tools for experimentation and algorithm verification.
Requirements
- Applied machine learning
- Signal processing, imaging, or an adjacent domain
- Research prototyping and production-quality system development
- Model evaluation, ablation, data leakage avoidance, and reproducibility
- Machine learning for physics-based or inverse problems
Responsibilities
- Build ML components for medical image reconstruction
- Define training and evaluation pipelines, datasets, and metrics
- Productionize models with performant inference, reproducibility, monitoring, and safe fallbacks
- Develop hybrid algorithms using physics and learned priors
- Build tools for rapid experimentation and rigorous algorithm verification