Winny Chepkemoi

Key Skills

Software

LabelImg

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Computer Code Programming

Document

Geospatial Tiled Imagery

Image

Medical Dicom

Top Label Types

Action Recognition

Audio Recording

Classification

Computer Programming Coding

Freelancer Overview

I am a detail-oriented professional with over five years of experience in quality assurance, data validation, and documentation review, primarily within healthcare settings. My background includes reviewing large volumes of sensitive records, meticulously checking for accuracy, completeness, and consistency, and following strict guidelines to ensure high-quality outputs. I am skilled in identifying errors, inconsistencies, and background noise in both written and audio content, and I regularly provide objective feedback to improve data quality. With strong written English skills, a keen eye for detail, and the ability to work independently in remote environments, I am well-equipped to handle data labeling, annotation, and AI training data tasks—especially those involving medical or clinical documentation. I am comfortable with repetitive review tasks, maintaining confidentiality, and meeting productivity targets while upholding the highest standards of accuracy.

Labeling Experience

AI Image annotator- metal object detection.

LabelimgImageBounding BoxEntity Ner Classification

This project involved labeling images of metal objects such as tools, machinery parts, and industrial equipment to create high-quality training datasets for AI models. Tasks included drawing bounding boxes around each object, assigning accurate classification labels (e.g., wrench, screwdriver, pipe), and ensuring consistency and precision across all images. The labeling was done using LabelImg, a widely-used annotation software. This dataset was designed to train AI systems for industrial inventory automation and object recognition.

This project involved labeling images of metal objects such as tools, machinery parts, and industrial equipment to create high-quality training datasets for AI models. Tasks included drawing bounding boxes around each object, assigning accurate classification labels (e.g., wrench, screwdriver, pipe), and ensuring consistency and precision across all images. The labeling was done using LabelImg, a widely-used annotation software. This dataset was designed to train AI systems for industrial inventory automation and object recognition.

2025 - 2025

Education

Work History