Warren Machuki

I am a detail-oriented Data Annotator and AI Trainer with over three years' experience contributing to production-level AI models across text, image, and audio domains. I have annotated and validated 15,000+ data points, consistently maintaining quality scores above 95% and often delivering 98% average accuracy by following complex, multi-layered guidelines.

I bring proven skills in image segmentation, bounding boxes, text categorization, speech transcription, search relevance evaluation, and conversational AI verification, with experience on platforms such as Appen and Toloka. I proactively participate in QA workflows, identify guideline ambiguities, and provide structured feedback to improve model performance and task clarity.