AI Data Labelling Contributor (Remotasks, Remote/Freelance)
Contributed to AI training data preparation by completing text classification and sentiment annotation tasks across thousands of data points. Labeled conversational AI responses for helpfulness, harmfulness, or being off-topic to support RLHF (Reinforcement Learning from Human Feedback) training pipelines. Reviewed and rated AI-generated content for factual accuracy, tone, and instruction-following quality to maintain consistency and audit-ready performance. • Text classification (helpful vs. harmful vs. off-topic) • Sentiment labeling • Factual accuracy and tone evaluation • Quality review against rubrics/benchmarks