Paul Njenga

I’m a results-driven Data Labeller & Annotation Specialist focused on clean, accurate, consistent training datasets for AI model development. I bring hands-on expertise across image, text, audio, and video annotation, applying complex guidelines with precision and speed.

In my current role as a Senior Data Annotation Specialist, I’ve annotated 500,000+ data points and consistently achieved a quality score of 97%+ across tasks. I perform bounding box, polygon, polyline, keypoint, semantic segmentation, NER, intent classification, and relation extraction for computer vision and NLP pipelines.

I also contribute to RLHF by evaluating and ranking AI-generated responses and providing detailed feedback for prompt quality assessment and model fine-tuning. I collaborate with project leads to resolve ambiguous labeling edge cases and refine annotation taxonomies.

Beyond production work, I mentor junior annotators and strengthen quality systems—reducing annotation error rates by 30% through internal QA checklists and feedback loops. I hold a BSc in Computer Science and complete training and certifications spanning machine learning, NLP, computer vision annotation tools, and data privacy.