Case Study: Lytx achieves scalable labeling and safer roads with Labelbox

Lytx Using Data to Help Save Lives on Our Roadways

Lytx, the market leader in video telematics and maker of the DriveCam safety program, uses video and ML to improve fleet safety across 500,000+ subscriptions and 850,000 drivers. As their labeling needs grew from a few thousand images to dozens of units collecting over 10,000 events per day, Lytx needed to scale review teams and infrastructure quickly and to monitor models in production. They turned to Labelbox for a scalable labeling and data infrastructure solution.

Labelbox provided a customizable labeling platform, custom interfaces and portals, reviewer management, and integration between labeling tasks and model predictions for QA and A/B testing. That infrastructure let Lytx rapidly scale review capacity to handle the high event volume, speed up labeling with templates, feed performance metrics into dashboards, and focus their team on modeling and evaluation—enabling production model monitoring and continued expansion of their ML initiatives with Labelbox.

Lytx

Stephen Krotosky

Applied Machine Learning