Case Study: Gamurs scales high-quality esports training data to build an AI coach with Labelbox

Gamurs - Customer Case Study

Gamurs, an AI-powered esports community platform, needed millions of specialized annotations to train an AI coach for games like Dota 2 and League of Legends. CTO Grant Osborne faced impractical costs from crowdsourcing and the expense of building an internal labeling team, so he turned to Labelbox’s managed outsourcing features and developer-focused API to centrally manage datasets, storage, and annotation workflows.

Labelbox recommended vetted BPO partners and hosted Gamurs’ internal and outsourced labelers in a single, transparent platform, enabling fast onboarding (about 4–5 weeks) and mixed-provider workflows for consensus QA. Gamurs deployed a 20-person and a 10-person labeling team (costs reported around 1.52¢ and 10–12¢ per bounding box respectively), scaled annotation throughput to meet millions of labels, and successfully productionized their ML pipeline while maintaining quality and operational oversight with Labelbox.

Gamurs

Grant Osborne

Chief Technology Officer