Outlier.ai — AI Data Specialist (AI output evaluation & dataset auditing)
Evaluated AI-generated outputs for precision, tone, relevance, and compliance with provided instructions to support high-quality model behavior. Applied detailed guidelines to assess language understanding, intent classification, and overall response quality for downstream use. Reviewed and corrected annotated datasets to maintain dataset consistency and fairness for training and QA pipelines. • Precision/tone/relevance assessment of generated responses • Intent classification and language understanding checks • QA auditing with error detection and hallucination spotting • Bias/ambiguity/ethical risk flagging with improvement feedback