Case Study: Swift Medical Scales High-Accuracy Wound Annotations with Sama

Swift Cost-Effectively Scaled Annotations Without Compromising Accuracy

Swift Medical, the company behind the Skin and Wound mobile app, needed a way to scale high-quality computer vision annotations for its AutoTissue model without sacrificing clinician-level accuracy. In-house labeling was too costly to expand, and other outsourced providers could not meet the quality bar, so the team turned to Sama for help.

Sama implemented specialized annotator training, continuous feedback loops with Swift’s clinical team, and a rigorous multi-step QA process to label wound regions and tissue types accurately. The partnership delivered strong measurable gains, including a 97% average internal quality score, a 230% increase in throughput from week 1 to week 10, and a 66.7% reduction in rejections over the same period.

Swift Medical

Dhanesh Ramachandram

Senior Machine Learning Researcher