Case Study: a sports AI company achieves faster, high-accuracy data labeling with Netsmartz

Sports Company tags 800 images per shift with Netsmartz

The Sports Company, an award-winning sports analytics AI startup, faced the challenge of tagging and labeling thousands of real-time images from sporting events with near 100% accuracy. They needed to identify athletes, jersey numbers, and gear using bounding boxes and segmentation to train their models, requiring a team with deep sports knowledge that could operate around the clock. Netsmartz provided a team of data annotators and a dedicated project manager to address this.

Netsmartz solved this by carefully selecting and training sports enthusiasts as annotators, creating a 24/7 shift system, and maintaining constant client communication. The results included annotating up to 800 images per shift and cleaning 2,000-3,000 bib images daily, achieving 95% accuracy within the first two weeks. This allowed the sports company to confidently expand into major team sports leagues with speed, accuracy, and cost efficiency.