Ambagwa E.

Key Skills

Software

Label Studio

Other

Top Subject Matter

Computer vision (images) dataset annotation for AI/ML training

Natural language processing (NER tagging) for AI/ML training

Top Data Types

Image

Text

Top Task Types

Bounding Box

Polygon

Segmentation

Classification

Entity (NER) Classification

Freelancer Overview

Personal AI Data Labeling Practice Project (Self-directed). Core strengths include Label Studio and Other. Education includes Diploma in Information Technology, Technical University of Kenya (2024). AI-training focus includes data types such as Image and Text and labeling workflows including Bounding Box, Polygon, and Segmentation.

Labeling Experience

Personal AI Data Labeling Practice Project (Self-directed)

Label StudioImageBounding BoxPolygon

Annotated an 800+ image self-curated dataset for computer vision training in a remote, self-directed workflow using a written custom guideline document. Applied multi-class labeling with bounding boxes and additional spatial label formats to ensure consistent dataset quality. Maintained inter-annotator consistency by running periodic self-review and comparing outputs against reference annotations from open-source datasets. • Used Label Studio for image annotation and structured exports. • Followed a custom multi-class annotation guideline with documented edge cases. • Verified accuracy and consistency through re-checking against reference labels. • Logged annotation decisions to build an audit-ready personal knowledge base.

2024 - Present

Text Classification & NER Tagging Experiment (Self-directed)

OtherTextEntity (NER) Classification

Performed text annotation for Named Entity Recognition by tagging 500+ sentences across Person, Organization, Location, and Date categories. Applied consistent labeling rules and resolved ambiguous cases using documented reasoning to support guideline adherence. Re-annotation review achieved 94% label consistency, demonstrating reliable performance over extended repetitive labeling. • Labeled entities using NER category taxonomy (Person, Organization, Location, Date). • Documented rationale for ambiguous span decisions. • Conducted review passes to measure and improve label consistency. • Supported downstream NLP training through structured, repeatable annotation.

2024 - 2024

Education

Work History