Data Annotation | Outlier AI (AI content evaluation and annotation)
Other
Text
TextPerformed evaluation and annotation of AI-generated content for accuracy, quality, and policy compliance. Reviewed large language model (LLM) outputs to detect problems in reasoning, tone, formatting, and instruction adherence. Contributed to iterative improvement by refining prompts and associated response data used for training and quality gains. • Annotated AI content against quality and compliance criteria • Assessed LLM responses for reasoning and instruction-following issues • Identified formatting and tone inconsistencies in generated text • Developed/refined prompts and response data to support model improvement
2025 - Present