Kelvin Wambua

I am an AI data specialist with a strong foundation in linguistics and over four years designing workflows, evaluation rubrics, and multilingual datasets for AI research and alignment. I focus on improving model performance, safety, and cultural relevance through structured, repeatable processes.

My hands-on experience includes producing high-accuracy annotations across text, image, and audio modalities, localizing Swahili datasets, and collaborating with global research teams to deliver datasets with 98%+ accuracy. I have a proven record of creating QA rubrics and documentation that scale across projects.

I hold a BA in Linguistics (First Class Honors) and have been recognized as a top contributor in multilingual AI training initiatives; I bring meticulous attention to detail, strong cross-cultural competence (Swahili–English), and practical expertise with tools like Label Studio, CVAT, Prodigy, and the Forte Platform.