LLM Response Evaluation & Rubric-Based Annotation (Legal & Business Domains)
Annotated and evaluated large language model outputs across legal and business-use cases using structured, multi-criteria rubrics. Tasks included factuality verification, logical reasoning assessment, compliance checks, and identification of subjective or unverifiable claims. Maintained high inter-annotator agreement through guideline refinement and conducted secondary QA reviews to ensure consistency, precision, and adherence to project standards.