Across freelance work on global AI platforms, I evaluated and benchmarked 500+ complex LLM prompts for reasoning, instruction-following, and safety. I also annotated video datasets for object detection and temporal segmentation, achieving an audited 98.5% precision score.
On a Swahili and regional-language LLM alignment project, I evaluated language, translated prompts, and supported cultural safety alignment for bilingual datasets. In hardware operations, I managed diagnostics and mobile repairs and authored SOPs and diagnostic guides.
