Technical AI Data Annotator & Domain Expert (Online Freelance / AI Training Platforms)
Reviewed, ranked, and evaluated LLM responses for technical accuracy, clarity, and safety as part of an RLHF-style training workflow. Wrote detailed written justifications for required model corrections to guide improved model behavior. Performed hallucination detection and ensured responses met domain-specific constraints for reliable training data. • RLHF response ranking and justification writing • Technical accuracy and safety evaluation • Hallucination identification and correction guidance • Clarification of edge cases in STEM responses