Ezekiel Nabea

Key Skills

Software

CVAT

Data Annotation Tech

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Image

Top Label Types

Bounding Box

Polygon

Freelancer Overview

I am an experienced AI Data Specialist with over 6 years of hands-on expertise in data annotation, labeling, and evaluation for AI training across text, image, audio, and video formats. My work focuses on both Swahili and English datasets, ensuring high-quality, culturally accurate, and linguistically sound data for machine learning and NLP applications. I am skilled in quality assurance, prompt and LLM testing, and compliance with strict guidelines, consistently maintaining over 95% quality accuracy. With a strong background in computer science and a proven track record in remote, collaborative environments, I am committed to delivering reliable, inclusive, and effective AI training data for diverse domains.

Labeling Experience

Swahili & English AI Data Annotation and Evaluation for LLM and Multimodal Systems

CVATImageBounding BoxPolygon

Worked as a Swahili and English AI Data Annotator and Evaluator across multiple remote projects supporting machine learning and large language model development. Responsibilities included annotating text, image, audio, and video datasets; evaluating AI-generated responses for accuracy, safety, fluency, and cultural relevance; performing pairwise comparisons and RLHF-style evaluations; and conducting linguistic quality assurance. Maintained consistent 95%+ accuracy while adhering to strict annotation guidelines and project-specific taxonomies. Contributed to improving AI model performance in multilingual and culturally sensitive environments.

Worked as a Swahili and English AI Data Annotator and Evaluator across multiple remote projects supporting machine learning and large language model development. Responsibilities included annotating text, image, audio, and video datasets; evaluating AI-generated responses for accuracy, safety, fluency, and cultural relevance; performing pairwise comparisons and RLHF-style evaluations; and conducting linguistic quality assurance. Maintained consistent 95%+ accuracy while adhering to strict annotation guidelines and project-specific taxonomies. Contributed to improving AI model performance in multilingual and culturally sensitive environments.

2020 - 2025

Education

Work History