AI Trainer (Mathematics & Generalist)
As an AI Trainer at Surge AI, I trained and evaluated large language models on mathematical reasoning and generalist tasks. My responsibilities included annotating and validating datasets, prompt engineering, outlier detection, and error analysis to ensure high-quality AI outputs. I provided structured feedback to enhance logic, accuracy, and consistency, improving the reliability of AI systems. • Annotated and validated datasets across algebra, calculus, and probability subjects. • Evaluated AI-generated responses for logical consistency and accuracy. • Designed and refined prompts to minimize hallucinations and boost reliability. • Conducted systematic outlier detection and comprehensive error analysis.