Clinical Image Segmentation ML Engineer — Production

Midjourney; San Francisco (CA); Hybrid; USD 140,000 - 190,000; Full time

Job summary

Midjourney is seeking an experienced ML engineer to own tissue-class segmentation and labeling models for ultrasound CT clinical analysis. You will design training pipelines, datasets, and evaluation metrics, while collaborating with clinicians and cloud teams to ensure robust, auditable model deployments.

Key tasks include retuning across 2D/3D inputs, defining dataset versioning, and ensuring compliant, low-latency inference with proper drift monitoring and fallbacks.

Qualifications

• Strong applied ML experience, architecting, training and evaluating models from scratch.
• Experience with image segmentation (2D/3D) and robust training-data choices.

Responsibilities

• Own tissue-class segmentation and labeling models and their training pipelines.
• Retune across 2D, 3D, and 2D×3D fusion as inputs update.
• Define pipelines, datasets, and metrics from ground up or open source.

Skills

Applied ML

Image segmentation

PyTorch

End-to-end model development

Design controls

Tools

MONAI

ITK / SimpleITK

3D Slicer

Job description

Midjourney is seeking an experienced ML engineer to own tissue-class segmentation and labeling models for ultrasound CT clinical analysis. You will design training pipelines, datasets, and evaluation metrics, while collaborating with clinicians and cloud teams to ensure robust, auditable model deployments.

Key tasks include retuning across 2D/3D inputs, defining dataset versioning, and ensuring compliant, low-latency inference with proper drift monitoring and fallbacks.