Bloomie S.

Key Skills

Software

Telus

Top Subject Matter

Artificial Intelligence / Machine Learning

Natural Language Processing (NLP)

Top Data Types

Text

Document

Top Task Types

Evaluation/Rating

Prompt + Response Writing (SFT)

Classification

Text Summarization

Text Generation

Object Detection

Question Answering

Freelancer Overview

In the last two years, I've been employed as an AI training data specialist for two top platforms Turing and Telus International. I was periodically engaged in assessing the outputs of LLM, annotating and labeling huge datasets, creating prompts, and generating augmented training data by paraphrasing and transforming texts. My areas of focus have included content classification, quality scoring, relevance ranking, ambiguity resolution, etc. My manner of working is always executing detailed rubrics and SOPs in a fully remote, high-volume context. I earned a BA in English and Literature from Stanford University, which has equipped me with a great knowledge of language, logic, and analysis that directly help me create accurate and dependable training data. Besides my linguistic knowledge, what differentiates me is my combination of skills, and hands-on experience with AI training. It's not just about labeling data for me I know the reason why quality is crucial at every stage of the pipeline and how my outputs impact model performance downstream. I am also skilled in working with LLMs like ChatGPT and Gemini, quickly changing with the new guidelines, and keeping precision while the project requirements are changing. I respect and learn from feedback, visibly getting better, and bring a high level of attentiveness to AI training tasks that have a great impact.

Labeling Experience

LLM Output Evaluation & Prompt Engineering

TextPrompt + Response Writing (SFT)

Worked as an AI Training Data Specialist for Turing, contributing to the development and quality improvement of large language models. Core responsibilities included evaluating and rating LLM-generated responses for accuracy, fluency, relevance, and instruction-following. Created and refined prompts designed to test and improve model reasoning and language capabilities. Produced paraphrased and rewritten text to expand training datasets, and classified content across multiple categories using detailed annotation rubrics. Maintained consistently high accuracy across large task volumes in a fully remote, flexible environment.

2025 - 2026

Education

Work History