Lee Atunga

Key Skills

Software

Clickworker

CVAT

Labelbox

Remotasks

Sama

Telus

Scale AI

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Video

Top Label Types

Bounding Box

Evaluation Rating

Translation Localization

Freelancer Overview

I have over five years of experience working on data labeling and AI training projects, mainly focused on computer vision and text annotation. At Samasource Kenya, I supported several large-scale AI initiatives from labeling street images for autonomous vehicle models to refining datasets for NLP systems, always maintaining accuracy rates above 98%. I am skilled in using tools like CVAT, Labelbox, VIA, and SRT HALO, and I take pride in delivering high-quality, well-structured data that helps machine learning models perform better. Beyond the technical work, I enjoy improving workflows and supporting team growth. I have trained and mentored new annotators, helped streamline annotation processes, and contributed to quality audits that boosted efficiency by 10%. I am detail-oriented, adaptable, and genuinely passionate about the role that accurate, human-labeled data plays in shaping reliable AI systems.

Labeling Experience

Text Intent and Sentiment Classification

Internal Proprietary ToolingVideoQuestion AnsweringText Summarization

Tagged and categorized text data to train NLP models for chatbots and sentiment analysis tools. This included verifying language context, tone, and intent to help improve the accuracy of conversational AI responses.

Tagged and categorized text data to train NLP models for chatbots and sentiment analysis tools. This included verifying language context, tone, and intent to help improve the accuracy of conversational AI responses.

2024 - 2025

Facial Landmark Detection

Don T DiscloseVideoPoint Key PointEntity Ner Classification

Used keypoint labeling to map facial features for AI models focused on emotion detection and face recognition. Maintained strict adherence to data privacy and ethical standards throughout the process.

Used keypoint labeling to map facial features for AI models focused on emotion detection and face recognition. Maintained strict adherence to data privacy and ethical standards throughout the process.

2020 - 2024

Autonomous Vehicle Image Annotation

SamaVideoBounding BoxPolygon

Labeled over 10,000 street and traffic images to train object detection models for self-driving car systems. My work involved drawing bounding boxes and polygons around road signs, vehicles, pedestrians, and lane markings to improve model accuracy in real-world driving conditions.

Labeled over 10,000 street and traffic images to train object detection models for self-driving car systems. My work involved drawing bounding boxes and polygons around road signs, vehicles, pedestrians, and lane markings to improve model accuracy in real-world driving conditions.

2017 - 2019

Education

Work History