Francis Kiarie

Key Skills

Software

CVAT

Top Subject Matter

Computer Vision

Historical Content Generation

Top Data Types

Image

Text

Top Task Types

Bounding Box

Freelancer Overview

I have experience in supporting AI training data workflows through evaluating, refining, and validating AI-generated content to ensure accuracy and consistency. I also design and test prompts, compare outputs across models, and use structured guidelines to identify and correct errors. In addition, I have basic hands-on experience with image annotation using CVAT, including bounding boxes, tagging, and classification. Further, my background in research and data analysis, using tools such as Excel, Google Sheets, and Tableau, has strengthened my ability to clean, analyze, and validate data, aiming to generate reliable and high-quality AI training data.

Labeling Experience

AI Research & Content Evaluation (Independent Project)

Text

Designed and executed an evaluation workflow to refine AI-generated historical text content. Crafted and iterated natural-language prompts, assessed LLM outputs for accuracy and clarity, and synthesized information into structured narratives. Applied constraints to AI outputs and conducted manual evaluations for factuality and coherence. • Reviewed and rated AI-generated text for hallucinations and factual accuracy. • Compared outputs across LLMs for tone, depth, and clarity evaluation. • Used prompt engineering to optimize AI text generation outcomes. • Enforced quality standards and consistency in labeled data before final content publication.

Designed and executed an evaluation workflow to refine AI-generated historical text content. Crafted and iterated natural-language prompts, assessed LLM outputs for accuracy and clarity, and synthesized information into structured narratives. Applied constraints to AI outputs and conducted manual evaluations for factuality and coherence. • Reviewed and rated AI-generated text for hallucinations and factual accuracy. • Compared outputs across LLMs for tone, depth, and clarity evaluation. • Used prompt engineering to optimize AI text generation outcomes. • Enforced quality standards and consistency in labeled data before final content publication.

2025 - Present

AI Data Annotation (Independent Practice)

CVATImageBounding Box

Performed image annotation tasks on sample datasets as part of independent practice in AI data annotation. Utilized the CVAT tool to create bounding boxes, perform image classification, and tag objects according to provided guidelines. Applied consistency rules, conducted quality checks, and achieved precision in object boundaries throughout the labeling process. • Labeled and reviewed images for bounding box accuracy and object detection quality. • Implemented and maintained strict adherence to annotation guidelines. • Conducted reviews to identify and correct mislabels and inconsistencies. • Practiced pixel-level accuracy and developed proficiency with annotation tools.

Performed image annotation tasks on sample datasets as part of independent practice in AI data annotation. Utilized the CVAT tool to create bounding boxes, perform image classification, and tag objects according to provided guidelines. Applied consistency rules, conducted quality checks, and achieved precision in object boundaries throughout the labeling process. • Labeled and reviewed images for bounding box accuracy and object detection quality. • Implemented and maintained strict adherence to annotation guidelines. • Conducted reviews to identify and correct mislabels and inconsistencies. • Practiced pixel-level accuracy and developed proficiency with annotation tools.

2026 - 2026

Education

Work History