Case Study: Yext boosts AI search labeling efficiency with HumanSignal's Label Studio

Yext boosts labeling capacity 250% with HumanSignal

Yext, a technology company providing an AI-powered search engine, needed to train its natural language processing models in multiple languages but was hindered by an inefficient and non-scalable data labeling tool from their previous vendor. The system's slow manual processes, lack of user visibility, and poor validation functionality caused significant data loss and threatened Yext's international expansion plans, leading them to seek a solution from HumanSignal's Label Studio Enterprise.

By implementing Label Studio Enterprise, HumanSignal provided Yext with a centralized platform that streamlined its annotation workflow. The solution enabled a more efficient review process, better team governance, and greater flexibility, resulting in annotators completing 2-4 times more work per day and a 250% increase in project capacity. The team also eliminated a 40% data discard rate, reduced the time to create training datasets by 1-2 weeks, and expanded the number of distinct annotation tasks it could perform.

Yext