I've worked as a freelance AI Data Specialist and AI Trainer, evaluating LLM-generated responses for truthfulness, instruction-following, safety, and helpfulness.
I annotate and label text datasets for machine learning pipelines with an average accuracy score of 98%, while crafting prompts that uncover edge cases and reduce output bias. I also fact-check AI-generated content against reliable online sources.
Previously, at a local Nairobi tech hub, I categorized, tagged, and cleaned unstructured text and visual data for client machine learning projects while maintaining strict annotation guidelines and delivery targets.
