Mindrift — AI Data Annotator / QA Reviewer (AI response evaluation, annotation QA, and search/product categorization)
Evaluated AI-generated responses for factuality, safety, clarity, and instruction adherence across multiple evaluation workflows. Performed pairwise comparison and ranking of search engine and LLM outputs to support model quality improvement and benchmarking. Reviewed annotated datasets and provided QA feedback to improve consistency, accuracy, and training readiness for contributors. • Response scoring and prompt review within AI benchmarking workflows • Pairwise ranking and comparison to identify more relevant/useful results • QA review of annotation outputs with correction feedback • Product categorization and data validation for AI training and recommendation systems