Cancer AI ML Engineer - Multimodal Biomarkers

ArteraAI; United States; Sur place; USD 140 000 - 180 000; Plein temps

Avantages offerts par ce poste

Equity compensation

401k matching

Unlimited PTO

Résumé du poste

Artera is seeking a machine learning engineer to help develop AI biomarkers that improve cancer care. You will work with ML scientists, engineers, and cross-functional partners across the model-development lifecycle, from prototyping through production deployment.

You will build and evaluate multimodal models, including pathology images and clinical data, focusing on robustness, interpretability, and documentation for regulatory quality processes.

Qualifications

• Experience developing ML/DL models with PyTorch or TensorFlow.
• Familiarity with oncology and biomarker development.
• Experience with real-world datasets and evaluation using appropriate metrics.

Responsabilités

• Develop and evaluate AI biomarkers from multimodal data to predict patient outcomes.
• Develop and evaluate self-supervised foundation models and downstream models (MIL, time-to-event, segmentation, classification).
• Improve model robustness across scanners, sites, staining protocols, and patient populations.
• Apply interpretability methods to explain model decisions and build clinician trust.
• Build and improve tools and workflows for reproducible model development, experimentation, validation, and deployment.
• Document experiments and communicate findings clearly; collaborate with cross-functional teams including regulatory partners.

Connaissances

Python programming

Collaboration

Experiment analysis

Communication of findings

Formation

Master's or graduate research

Outils

PyTorch

TensorFlow

Description du poste

Artera is seeking a machine learning engineer to help develop AI biomarkers that improve cancer care. You will work with ML scientists, engineers, and cross-functional partners across the model-development lifecycle, from prototyping through production deployment.

You will build and evaluate multimodal models, including pathology images and clinical data, focusing on robustness, interpretability, and documentation for regulatory quality processes.