Simon N.

Key Skills

Software

Remotasks

Labelbox

CVAT

Appen

Micro1

Top Subject Matter

Technology - AI training and Data annotoation

Data - Text labelling and Science

Transcription - Video and Audio

Top Data Types

Image

Text

Video

Top Task Types

Bounding Box

Data Collection

Classification

Text Generation

Fine-tuning

Transcription

Prompt + Response Writing (SFT)

Freelancer Overview

Over the past year, I have worked directly in AI training and data operations, most notably as a Volunteer AI Data Trainer at Obondium, where I designed structured instructional content and evaluated AI-generated outputs for correctness, policy compliance, and reasoning quality. I iteratively refined prompts and training materials based on evaluation findings, documented edge cases and decision rationales, and applied content safety principles when curating data for AI pipelines. Alongside this, my academic and project work involved building and evaluating machine learning models end-to-end - from data acquisition and cleaning through training, evaluation, and deployment - giving me a strong understanding of what quality training data looks like and why it matters.

Labeling Experience

AI Training Data & Prompt Evaluation — RWS, Outlier & Obondium

TextClassification

Accumulated hands-on AI training data experience across multiple platforms and organizations including RWS, Outlier, and Obondium. Work has spanned prompt engineering, response evaluation, data annotation, and quality review for large language model (LLM) training pipelines. Responsibilities included writing and refining structured prompts to improve AI output quality, rating and evaluating model-generated responses for accuracy, helpfulness, reasoning quality, and policy compliance, and annotating text data according to detailed labeling guidelines. Applied content safety awareness when flagging harmful, ambiguous, or non-compliant outputs, and documented edge cases and decision rationales to support consistent labeling across teams. Supported diverse task types across multilingual and multi-locale contexts, with strong attention to nuance and guideline adherence.

2024 - 2026

Education

Work History