Matthew O.

Key Skills

Software

Data Annotation Tech

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Document

Image

Text

Top Task Types

Classification

Evaluation/Rating

Prompt + Response Writing (SFT)

Question Answering

RLHF

Freelancer Overview

I have extensive experience in data labeling and AI training data evaluation, particularly in assessing large language models (LLMs) using structured methodologies. At Outlier.ai, I contributed to projects such as Flamingo, Goldfish Crackers, and Nightingale, where I conducted side-by-side (SxS) evaluations and systematically rated model outputs using a Likert scale. This approach allowed for a precise assessment of response quality, accuracy, completeness, and adherence to prompt instructions. My expertise lies in identifying subtle distinctions in AI-generated content and providing structured feedback to enhance model reliability and contextual relevance. Through my work, I have developed strong analytical skills in evaluating diverse datasets, refining AI systems to improve their overall performance and alignment with user expectations.

Education

Work History