Shira Kimble

Key Skills

Software

Telus

Top Subject Matter

No subject matter listed

Top Data Types

Video

Top Label Types

Bounding Box

Freelancer Overview

I am a detail-oriented data annotation and QA specialist with over six years of experience supporting AI training systems through precise image-based data labeling, text extraction, and answer validation. My background in English and professional writing enables me to deliver native-level fluency, semantic accuracy, and clear, standalone responses for machine learning projects. I have a proven track record in reviewing and correcting image-based Q&A datasets, verifying label accuracy, and providing structured feedback to improve annotation quality. Skilled in tools like Labelbox, Scale AI, Appen Connect, Jira, and Airtable, I excel in remote, deadline-driven environments that demand consistency, critical thinking, and rigorous data quality assurance. My contributions have been recognized for reducing labeling errors and maintaining top-tier accuracy benchmarks across computer vision and NLP tasks.

Labeling Experience

Image-Based QA Annotation & Linguistic Evaluation for AI Training

TelusVideoBounding Box

Evaluated and approved large-scale image-based question-answering datasets for training multimodal AI models. Duties included checking the correctness of labels, evaluating the correctness of answers, and checking for consistency between images and answers. Identified accurate text from images for evidence-based validation and reformatted answers for self-contained clarity, grammatical correctness, and fluent natural U.S. English language usage. Used structured evaluation criteria for classification, question-answering, and text generation tasks, maintaining quality scores above 98% at all times. Identified ambiguous or unanswered questions with sound reasoning and provided actionable feedback to annotators to minimize systematic errors. Performed linguistic evaluation for grammar, punctuation, semantic correctness, tone, and relevance. Completed work on various platforms (Labelbox, Scale AI, Appen Connect, TELUS AI) in a fast-paced remote setting, averaging about 7 minutes per task

Evaluated and approved large-scale image-based question-answering datasets for training multimodal AI models. Duties included checking the correctness of labels, evaluating the correctness of answers, and checking for consistency between images and answers. Identified accurate text from images for evidence-based validation and reformatted answers for self-contained clarity, grammatical correctness, and fluent natural U.S. English language usage. Used structured evaluation criteria for classification, question-answering, and text generation tasks, maintaining quality scores above 98% at all times. Identified ambiguous or unanswered questions with sound reasoning and provided actionable feedback to annotators to minimize systematic errors. Performed linguistic evaluation for grammar, punctuation, semantic correctness, tone, and relevance. Completed work on various platforms (Labelbox, Scale AI, Appen Connect, TELUS AI) in a fast-paced remote setting, averaging about 7 minutes per task

2019 - 2024

Education

Work History