AI Trainer / Labeler for Civic Space Violation Tracking AI
Built an agentic AI system that tracked 53 instances of civic space or freedom of expression violations in West Africa between April and June 2024. The process included identifying, classifying, and documenting relevant textual reports for AI training and tracking purposes. The AI training workflow primarily revolved around labeling and classifying various forms of civic space violations for accurate detection in the dataset. • Labeled and classified reports or articles of civic space violations for AI use. • Maintained structured datasets to ensure accurate tracking of labeled events. • Coordinated with NGOs to verify and refine labeled data samples. • Supported the AI development pipeline through regular data annotation and validation tasks.