Contact
LOG IN
Sign up

How RadAI Uses Weights & Biases to Ship Production ML Models Faster

"The pipeline we built, which has W&B registry as its first step, allowed us to reduce the need for synchronization across researchers and software engineers on the ML organization—less meetings, less documents."
Hasan Ali Demirci
Staff Software Engineer, Machine Learning


RadAI is on a mission to help radiologists focus on what matters most: solving diagnostic puzzles and improving patient care. Founded in 2018 by Dr. Jeff Chang, the youngest radiologist in U.S. history, RadAI builds AI-native tools like RadAI Impressions and RadAI Reporting that reduce burnout and give radiologists time back in their day.

In this video, Hasan Ali Demirci from RadAI’s ML team shares how the team uses Weights & Biases to move models from research to production. By using W&B’s model registry to trigger CI/CD pipelines, researchers can tag an artifact and kick off an end-to-end deployment, cutting down on meetings and cross-team overhead.

Today, every production model at RadAI is tracked and deployed through Weights & Biases, a core part of how the team ships reliable AI at scale.

Request demo