Ray Keicy

I’m a detail-oriented, analytical recent graduate focused on high-quality AI/ML training data. I’m experienced in image & video annotation (including bounding boxes and polygons), as well as text classification, sentiment labeling, and audio transcription tasks.

As a Freelance Data Annotator with Appen/Remotasks, I’ve completed 500+ annotation tasks across image classification, NLP, and audio transcription, consistently maintaining accuracy above 97%. I apply detailed labeling guidelines for object detection datasets and proactively flag ambiguous or low-quality sources to help reduce downstream model errors.

Before annotation work, I supported data integrity as a Data Entry & Research Assistant intern at the Kenya National Bureau of Statistics, entering and validating 2,000+ household survey responses with <0.5% error rate. I also strengthened my analytical foundation through coursework and a final-year project building and evaluating a sentiment-labeled Swahili-English code-switching Twitter dataset.