AI Training Specialist, Data Annotator and Reviewer
I conducted large language model evaluations using rubric-based side-by-side scoring and backend debug info analysis for advanced LLM personalization. My responsibilities included providing grounded rationales for RLHF, verifying user intent capture, and ensuring safe, accurate outputs. I labeled and evaluated high volumes of multilingual text, audio, and video data primarily supporting natural language processing for Indonesian and international contexts. • Audited LLM performance across SxS comparative assessments and backend data analysis. • Executed labeling and evaluation for multilingual audio and video text data, focusing on local language fidelity. • Implemented annotation and evaluation methods to support prompt refinement and model localization. • Assessed and rated website search results and content, contributing structured enhancement feedback for annotation guidelines.