AI Language Trainer & Content Evaluator — Outlier AI (Remote)
Performed structured AI evaluation for Aether Generalist Tasks, including RLHF and instruction-following style assessments. Rated outputs for factual accuracy, tonal alignment, and instruction adherence using standardized evaluation rubrics. Identified issues such as hallucinations, logical inconsistencies, and coherence failures across hundreds of consecutive sessions. • Delivered 6+ hours of daily evaluation at sustained 4.0+/5.0 quality ratings. • Covered multiple modalities including RLHF, instruction-following, Omni ELO, and text-to-image annotation. • Assessed responses for quality, compliance with prompts, and overall coherence. • Maintained platform quality across a high volume of evaluation sessions.