As a freelance AI Data Evaluator and LLM Response Reviewer, I rate and rank model responses against client rubrics, fact-check outputs, and flag hallucinations, bias, and unsafe content. I evaluate fluency and cultural accuracy in Yoruba, English, Korean, and Nigerian Pidgin.
In my multilingual LLM evaluation portfolio, I built a four-dimension rubric and applied it to model responses, documenting failure patterns such as missing Yoruba tone marks and imported cultural assumptions. I also created annotation guidelines and a six-field label schema for a self-initiated Pidgin and Yoruba dataset project.
As a freelance Executive Virtual Assistant, I manage calendars, inboxes, and travel across time zones, and prepare emails, meeting notes, reports, and presentations. I coordinate meetings and follow-ups using Google Workspace, Notion, and Trello.

