DataAnnotation — AI Trainer (Generalist)
Served as an AI Trainer for Generalist LLM improvement by rating, evaluating, comparing, and revising thousands of model responses to user-generated prompts. Assessed outputs across multiple dimensions including instruction following, truthfulness, completeness, writing style, and tone. Contributed to defining what high-quality answers look like by generating rubrics and criteria tied to specific prompts. • Rated response quality against prompt requirements and evaluation criteria • Compared multiple responses and suggested revisions to improve correctness and helpfulness • Evaluated across modalities including photo, audio, video, and text prompt/response modes • Produced rubrics/criteria to standardize ideal response expectations