Freelance Data Analyst / AI Evaluator (Remote)
TextFreelance work involved evaluating AI-generated responses for factual accuracy, logical consistency, and overall relevance. It also included performing annotation and classification tasks on text datasets to support training and quality improvement. The role focused on detecting low-quality or inconsistent outputs and correcting labeled data while reviewing prompts to improve system reliability. • Evaluated responses using factual correctness, logic, and relevance criteria • Annotated and classified items in text datasets for downstream training • Identified hallucinations/low-quality outputs and resolved inconsistencies in labels • Conducted research and verification to validate training dataset content
2020 - Present