AI Writer and Evaluator, Contract
As an AI Writer and Evaluator, I audited retrieval-augmented generation (RAG) outputs for accuracy, quality, and grounding in reference texts. I evaluated team content, provided actionable feedback, revised responses using domain expertise, and crafted complex, image-based prompts to test model reasoning. I contributed to supervised fine-tuning (SFT), rated and ranked outputs per reinforcement learning from human feedback (RLHF), and annotated intricate images with comprehensive descriptions in limited timeframes. • Assessed and refined both text and image AI-generated content. • Composed prompts targeting multimodal and general instruction-following abilities. • Provided feedback for team improvement and SFT processes. • Created high-quality annotations under strict time constraints.