LLM Content Evaluator (Kiswahili) - Sama
You evaluate and improve AI-generated and human-written Kiswahili content for generative AI training and benchmark readiness. You assess outputs for factuality, completeness, brevity, and grammatical correctness while refining rubrics to reduce evaluation inconsistencies. You also generate original Kiswahili entries from prompts and apply AI/ML fundamentals to strengthen annotation workflows. • Annotated 700+ Kiswahili datasets for content quality assessment • Reviewed AI outputs and human samples for alignment and coherence • Produced prompt-based original Kiswahili data with research support • Collaborated to update rubrics and support fine-tuning use cases