Kelvin Nyahe

Key Skills

Software

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Image

Top Label Types

Bounding Box

Cuboid

Tracking

Routing

Mapping

Freelancer Overview

I am a detail-oriented AI Data Annotation Specialist with over four years of experience supporting AI training data projects and structured data analysis. My expertise includes interpreting SQL query outputs, creating natural-language question-and-answer pairs, and performing rigorous data quality assurance to ensure consistency and accuracy. I have a strong background in remote, project-based environments, collaborating with AI training teams to refine annotation guidelines and enhance dataset quality. My technical toolkit includes SQL (MySQL, PostgreSQL), Microsoft Excel, Google Sheets, Salesforce, HubSpot, and basic Python (Pandas), and I am comfortable using remote collaboration tools like Slack, Microsoft Teams, and Jira. With a solid foundation in account management and CRM systems, I bring a unique perspective to data annotation tasks, ensuring that both business logic and linguistic clarity are maintained throughout every project. I am available 10–15 hours per week and am committed to delivering high-quality results in AI data labeling and annotation roles.

Labeling Experience

Image Annotation Specialist – E-commerce Product Catalog

LabelboxImageBounding BoxCuboid

Worked on a large-scale image annotation project for an e-commerce platform, labeling over 50,000 product images for accurate object detection and classification. Responsibilities included defining bounding boxes, categorizing products into specific categories, and ensuring labeling consistency across datasets. Applied quality control protocols to maintain labeling accuracy above 98%. Leveraged Labelbox to streamline workflow and collaborate with a remote annotation team, improving dataset usability for machine learning model training in visual search and recommendation systems.

Worked on a large-scale image annotation project for an e-commerce platform, labeling over 50,000 product images for accurate object detection and classification. Responsibilities included defining bounding boxes, categorizing products into specific categories, and ensuring labeling consistency across datasets. Applied quality control protocols to maintain labeling accuracy above 98%. Leveraged Labelbox to streamline workflow and collaborate with a remote annotation team, improving dataset usability for machine learning model training in visual search and recommendation systems.

2022 - 2023

Education

Work History