Data Analyst — Afritive AI Solutions (Jan 2025 – Present)
TextDesigned and validated large-scale data pipelines and quality control processes that mirror annotation QA used for LLM training and evaluation. Performed rigorous data validation and reconciliation by applying rule-based checks and human judgment to flag inconsistent or ambiguous entries for training readiness. Authored and enforced labeling guidelines and data documentation standards to ensure consistent, reproducible dataset outputs. • Analyzed datasets to detect outliers and edge cases relevant to evaluating LLM correctness and safety. • Translated complex requirements into actionable data tasks in collaboration with engineers and product stakeholders. • Produced structured documentation and evaluation-oriented guidelines for cross-functional teams. • Implemented data quality controls using Kafka and Python workflows transferable to annotation QA.