As an independent contractor, I evaluate and score LLM-generated responses across summarization, reasoning, and creative generation against quality parameters.
I identify hallucinations, factual errors, logical inconsistencies, and safety policy violations, and document my reasoning clearly. My evaluations follow RLHF protocols and assess safety, contextual accuracy, and prompt alignment.
I also annotate image and audio datasets for computer vision and voice recognition models, with accuracy ratings above 98%. I follow multi-page annotation rubrics in rapid-turnaround workflows.
On global freelance platforms, I work as an Online English Conversation Partner and Proofreader. I help learners develop oral fluency and proofread English transcripts, audio clips, and digital documents.
