Paul Musembi

I am an AI Training Specialist with over three years of hands-on experience in data labeling, annotation, and dataset preparation for machine learning models. I specialize in image, video and audio annotation and have strong familiarity with industry-standard tools and formats.

I have annotated large-scale datasets for object detection, tracking, semantic and instance segmentation, and consistently maintained annotation quality above 95%. I work with YOLO formats and frameworks, Labelbox, CVAT, Roboflow and VGG Image Annotator.

I collaborate closely with ML engineers to prepare training, validation and testing datasets, review and correct team annotations, and contribute to model improvement cycles. I have experience on surveillance multi-object tracking and real-time object detection projects.

I bring attention to detail, strong communication and a focus on data security and quality assurance to every project, and I aim to deliver datasets that accelerate reliable model training and deployment.