Gideon A.

Key Skills

Software

Appen

Clickworker

CrowdSource

Google Cloud Vertex AI

Labelbox

Label Studio

Mercor

Mindrift

Roboflow

Telus

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Data Collection

Evaluation/Rating

Prompt + Response Writing (SFT)

Text Generation

Text Summarization

Freelancer Overview

I am a product designer with a strong foundation in user research, usability, problem framing, and customer experience, which naturally expanded into AI data labeling and model training work. My background in UI design, product strategy, and real user interaction gives me the ability to understand context deeply when annotating text, audio, image or conversational data, because I can identify user intent, clarity, sentiment, and the subtle differences in meaning that influence model behavior. I have worked as a data annotator across multiple platforms including Outlier, Prolific, CloudGen, Oneforma and Telus where I contributed to training datasets for NLP, content evaluation, sentiment reviews, search relevance, and AI performance improvement tasks. I enjoy working precisely, following structured labeling guidelines, and providing consistent high-quality feedback that supports model alignment and accuracy. I am detail-oriented, able to adapt quickly to new task instructions, and committed to contributing meaningful labeling that improves AI reliability over time.

Labeling Experience

LLM Evaluation and Conversational Text Annotation Project

AppenTextClassificationQuestion Answering

This project involved reviewing AI generated responses, human conversational text, prompts and model outputs to evaluate clarity, correctness, coherence, safety and relevance. I performed content classification, intent identification, sentiment interpretation, rewriting/improvement of model responses and quality scoring based on task instructions. The project required strict guideline adherence, detailed reasoning consistency, and maintaining high accuracy across multiple annotation batches. I mainly worked on English data and contributed to improving reasoning depth, factual alignment, and user experience quality in LLM interactions. Approx project size: Over 250+ tasks completed across multiple batches.

This project involved reviewing AI generated responses, human conversational text, prompts and model outputs to evaluate clarity, correctness, coherence, safety and relevance. I performed content classification, intent identification, sentiment interpretation, rewriting/improvement of model responses and quality scoring based on task instructions. The project required strict guideline adherence, detailed reasoning consistency, and maintaining high accuracy across multiple annotation batches. I mainly worked on English data and contributed to improving reasoning depth, factual alignment, and user experience quality in LLM interactions. Approx project size: Over 250+ tasks completed across multiple batches.

2025 - 2025

Audio and Video Content Annotation for Speech Intent and Scene Understanding

TelusVideoClassificationEmotion Recognition

This project involved annotating short video clips and audio samples to identify speaker intent, classify actions, detect visible objects, and evaluate emotional tone. I performed labeling that focused on human conversation context, scene interpretation, sentiment, and behavior recognition to support training multimodal AI models capable of understanding both sound and visual data. I ensured label precision using task guidelines and consistent criteria to maintain annotation quality across large volumes of samples. Worked primarily in English language content, contributing over 180+ completed tasks across multiple annotation batches.

This project involved annotating short video clips and audio samples to identify speaker intent, classify actions, detect visible objects, and evaluate emotional tone. I performed labeling that focused on human conversation context, scene interpretation, sentiment, and behavior recognition to support training multimodal AI models capable of understanding both sound and visual data. I ensured label precision using task guidelines and consistent criteria to maintain annotation quality across large volumes of samples. Worked primarily in English language content, contributing over 180+ completed tasks across multiple annotation batches.

2023 - 2023

Education

Work History