AI Data Trainer & Quality Analyst at Scale AI (remote)
Trained and evaluated large language models by labeling over 2,500 text data points daily with 99.2% accuracy. Performed sentiment analysis, entity recognition, and prompt response scoring to improve model behavior. Reduced error rates by 35% through consistent quality assurance and iterative guideline refinement.• Labeled and scored model outputs for sentiment, entities, and response quality• Performed prompt-response evaluation to support model training and benchmarking• Maintained top-tier annotation quality (top 5% and repeated Star Performer awards)• Collaborated with global teams on training guideline updates and onboarding