At DigitalDivide Data (DDD Kenya), I annotate LiDAR and 3D data for autonomous-vehicle training, identifying people, vehicles, road signs, and other relevant objects. I track when vehicle blinkers and special-vehicle lights switch on and off, maintaining 99.95% average annotation accuracy.
Also at DigitalDivide Data (DDD Kenya), I annotate cocoa pods in tree images for agricultural computer-vision datasets, classifying them as ripe, non-ripe, or unknown. I review object boundaries and labels against project guidelines and flag ambiguous examples.
At SamaSource, I analyzed satellite imagery for geospatial datasets and reviewed AI-generated descriptions against visual evidence. I later reviewed other associates’ outputs in a quality assurance workflow; in freelance annotation work, I labeled datasets for AI moderation models and maintained 95%+ annotation accuracy across quality-reviewed projects.
