Labelbox
51 Case Studies
A Labelbox Case Study
Advent Health Partners, a healthcare service and technology company, faced the challenge of extracting data from millions of unstructured medical records. These documents, often hundreds of pages long and arriving in various formats, were time-consuming and expensive to process for their clinical ML models. They turned to the Labelbox platform to address this.
Labelbox provided a solution using active learning and model-assisted workflows. This enabled efficient human review and created a page-classification model to auto-tag records. The result was a dramatic speed increase in processing: the time to classify and review each page dropped from 13 seconds to 8 seconds. This cut 25 hours from the process of generating high-quality AI data, according to Advent Health Partners.