AI Trainer — March 2026 – Present
Text
Evaluation/RatingAI Trainer responsible for evaluating AI-generated responses from large language models for accuracy, relevance, helpfulness, and instruction compliance. You compared multiple candidate model outputs and selected preferred responses using established quality guidelines. You performed data annotation and response rating to support RLHF-style quality improvement for English and Indonesian content. • Response rating using rubrics and quality guidelines • Instruction-following and compliance assessment • Hallucination, factual inconsistency, and logical error checking • Language evaluation for English-Indonesian outputs
2026 - Present