Case Study: Stanford Medicine achieves 50K+ images labeled in minutes, replacing person‑years of manual labeling with Snorkel AI

Stanford Medicine - Customer Case Study

Stanford Medicine researchers faced a costly, time‑consuming challenge: labeling medical imaging and monitoring data for triage models required person‑months to person‑years of radiologist time. They engaged Snorkel AI and its cross‑modal Snorkel pipeline to accelerate the creation of training labels.

Snorkel AI’s solution generated high‑quality labels in hours, replacing extensive manual labeling—labeling 50,000+ images in minutes while achieving about 94% ROC AUC—and is currently being tested for deployment across Stanford and the Department of Veterans Affairs hospital systems.

Stanford Medicine