Generalist AI Trainer — Outlier.ai (AI training/evaluation)
Evaluated AI model outputs against detailed technical guidelines to identify and document failure patterns for engineering review. Performed quality-focused assessment work in a fully remote, results-driven setup. Contributed structured observations intended to improve model behavior and reduce recurring errors. • Reviewed model outputs and compared them to expected criteria • Logged failure patterns and common error modes in documentation • Ensured assessments met agreed quality and throughput standards • Prepared findings for engineering analysis and iteration