Lewis Macharia

Key Skills

Software

Clickworker

Top Subject Matter

No subject matter listed

Top Data Types

Video

Top Label Types

Bounding Box

Polyline

Freelancer Overview

I am a dedicated Data Labeling & AI Training Specialist with over a year of hands-on experience annotating images and videos for machine learning applications, particularly in computer vision domains such as object detection, action recognition, and tier-based labeling. My work has involved segmenting thousands of video clips, applying bounding boxes and polygons, and maintaining strict compliance with complex guideline systems like OpenTrain. I consistently achieve 98–99% QA accuracy while meeting high-volume production targets and am skilled at identifying edge cases and ensuring dataset reliability. I am proficient with tools such as CVAT, Labelbox, Supervisely, and use Google Sheets and Excel for tracking and quality assurance. My strong attention to detail, rapid learning of new labeling rules, and commitment to data quality make me a valuable contributor to any AI training data project.

Labeling Experience

Video Bounding Box Annotation for Human Activity Recognition

ClickworkerVideoBounding BoxPolyline

Annotated large-scale video datasets using bounding boxes to track human movements and object interactions for computer vision models. Performed frame-by-frame labeling of actions such as walking, reaching, holding, and placing objects while following strict annotation guidelines. Handled edge cases including partial occlusions and fast motion to maintain labeling accuracy. Completed hundreds of video segments with consistent quality checks, achieving high precision and meeting project turnaround deadlines for AI training workflows.

Annotated large-scale video datasets using bounding boxes to track human movements and object interactions for computer vision models. Performed frame-by-frame labeling of actions such as walking, reaching, holding, and placing objects while following strict annotation guidelines. Handled edge cases including partial occlusions and fast motion to maintain labeling accuracy. Completed hundreds of video segments with consistent quality checks, achieving high precision and meeting project turnaround deadlines for AI training workflows.

2025 - 2025

Education

Work History