AI Data Specialist & Linguistic Evaluator (AI evaluation, pairwise comparison, alignment scoring)
Provided AI alignment and linguistic evaluation for Oromo-language generative outputs, focusing on safety and helpfulness checks. Conducted pairwise structural comparisons to determine integrity and cultural safety of model responses. Performed expert reviews of reasoning paths to surface logic failures and linguistic hallucinations in generated text.• Reviewed generative Oromo responses using side-by-side pairwise comparison protocols. • Rated or flagged outputs based on linguistic correctness and safety criteria. • Documented issues identified during quality assurance audits. • Supported iterative improvement of Oromo content quality via evaluator feedback.