Amos Abuya

Key Skills

Software

CVAT

Dataloop

Labelbox

Sama

Toloka

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Label Types

Bounding Box

Classification

Polygon

Prompt Response Writing SFT

Segmentation

Freelancer Overview

I am a detail-oriented Data Annotation Specialist with over three years of experience supporting AI and machine learning projects, primarily through large-scale data labeling, tagging, and validation. My background includes working with image and text datasets, where I consistently follow strict annotation guidelines to ensure high-quality training data for AI models. I am skilled in using tools like Microsoft Excel and Google Sheets, and I am comfortable handling repetitive data tasks with accuracy and efficiency. My experience at Samasource has strengthened my ability to meet tight deadlines, maintain productivity, and collaborate effectively in remote or hybrid environments. I am dedicated to delivering precise, high-quality datasets that drive successful AI outcomes.

Labeling Experience

Annotation Project

LabelboxImageBounding Box

The project involved bounding box annotation for retail shelf object detection, aimed at training computer vision models for product recognition and inventory analysis. I labeled individual consumer goods by drawing precise bounding boxes around products on crowded supermarket shelves, including overlapping and partially occluded items. The work covered large batches of high-resolution images with multiple objects per image. Strict annotation guidelines were followed to ensure accuracy, consistency, and minimal background inclusion, with self-review and quality checks applied to meet defined QA and productivity standards.

The project involved bounding box annotation for retail shelf object detection, aimed at training computer vision models for product recognition and inventory analysis. I labeled individual consumer goods by drawing precise bounding boxes around products on crowded supermarket shelves, including overlapping and partially occluded items. The work covered large batches of high-resolution images with multiple objects per image. Strict annotation guidelines were followed to ensure accuracy, consistency, and minimal background inclusion, with self-review and quality checks applied to meet defined QA and productivity standards.

2025 - 2025

Education

Work History