Azure AI Foundry

#Overview

Azure AI Foundry is Microsoft's model catalog and deployment surface on Azure, where biological and healthcare foundation models deploy to managed compute endpoints for hosted inference. Its Health & Life Sciences filter surfaces first-party and partner models focused on the medical-imaging, radiology-reporting, and digital-pathology tasks that protein-centric hubs skip. For teams already building on Azure, it is a direct path to running clinical imaging inference behind enterprise identity, networking, and compliance controls.

#What you can run on Azure AI Foundry

MAIRA-2, Microsoft's multimodal radiology model, is served in the catalog as CXRReportGen for grounded chest X-ray report generation, turning radiographs into structured findings. Prov-GigaPath, a whole-slide digital-pathology foundation model offered as a partner listing, produces tile- and slide-level representations for computational pathology. Between them they cover medical imaging, radiology report generation, and histopathology inference, the slice of the catalog most relevant to diagnostics and imaging teams.

#Inference and fine-tuning on Azure AI Foundry

Models deploy one-click to Azure Machine Learning managed compute endpoints, giving you hosted APIs without standing up your own inference stack. The broader Foundry catalog supports fine-tuning many models on your own data, and everything runs under Azure's enterprise controls for identity, networking, and compliance rather than raw weight downloads. It suits hospital, diagnostics, and life-sciences teams that need governed, in-cloud deployment of imaging and pathology models with the scaling and security Azure already provides.