Katalina Tapia

AI content evaluator and prompt engineer with two years of experience building evaluation prompts, ranking model outputs, and annotating data for RLHF pipelines.

Background in Information Technology gives a strong basis for judging tone, clarity, and instruction-following in model responses. Comfortable working independently across shifting rubrics and large data volumes.

AI Content Reviewer & Writer - Scale AI - Remote

(2023-01)

• Wrote evaluation prompts, including edge cases, to test model behavior at its limits
• Ranked AI-generated responses against rubrics covering helpfulness, accuracy, depth, and safety
• Ran side-by-side comparisons of multiple model outputs and documented the reasoning behind each choice
• Flagged hallucinations, reasoning gaps, and policy violations in model outputs
• Annotated data for intent, tone/sentiment, helpfulness, and safety compliance
• Wrote structured feedback identifying model weaknesses and areas for improvement
• Maintained consistency across datasets while following evolving project guidelines

Data Annotation Specialist (Independent) - Mercor.AI - Remote

(2024-01 - 2025-01)

• Produced written content across technology, AI, and education topics
• Edited drafts for clarity, tone consistency, and grammar
• Simplified technical concepts into plain-language writing for general audiences

Bachelor of Science - Information Technology - University of Maryland Global Campus