Case Study: American Family Insurance achieves automated claims labeling with Labelbox

Labelbox Adapts to Support American Family Insurance Automation

American Family Insurance, a Fortune 400 insurer, needed to automate claims processing with machine learning but required a richer labeling ontology than their tools supported. They partnered with Labelbox and used Labelbox’s data-labeling platform (configurator and image-segmentation interface) to address the need for hierarchical, taxonomical labels that go beyond a single nesting level.

Labelbox implemented a nested dropdown/nested classification feature that supports arbitrary-depth taxonomies, letting admins define hierarchies in the configurator that the image-segmentation interface consumes. This change enabled American Family Insurance to train models using a reusable, hierarchical ontology—accelerating claims automation, avoiding costly in-house labeling infrastructure, and reducing engineering time and expense while standardizing the labeling workflow.

American Family Insurance