Owen Joseph

Key Skills

Software

Data Annotation Tech

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Image

Text

Video

Top Task Types

Action Recognition

Audio Recording

Bounding Box

Classification

Data Collection

Entity Ner Classification

Evaluation Rating

Point Key Point

Prompt Response Writing SFT

Relationship

Segmentation

Translation Localization

Freelancer Overview

I am a detail-oriented AI analyst with hands-on experience in data labeling, annotation, and AI model evaluation, particularly within decentralized AI, Web3, and predictive modeling domains. My work involves generating structured training content, designing prompts to test large language models, and assessing outputs for logical consistency, bias, and hallucinations. I have developed frameworks to evaluate AI responses for accuracy and quality, and I am skilled in using tools such as ChatGPT, Google Sheets, Excel, and basic Python for data manipulation. My strong analytical and research skills, combined with a keen eye for detail, enable me to produce high-quality, reliable training data that enhances AI system performance.

Labeling Experience

AI Content Analyst

Data Annotation TechVideoPoint Key PointEntity Ner Classification

In this project, I worked on training a natural language processing model to classify customer support queries, with the goal of enabling the AI to automatically categorize requests and suggest relevant responses. My role focused on annotating and labeling thousands of tickets, identifying intents and entities, normalizing text, and ensuring consistent application of labeling guidelines. The dataset included over 50,000 entries across multiple languages and product lines, and I personally annotated around 15,000, collaborating closely with a team of four to ensure coverage. To maintain high quality, we followed strict annotation protocols, performed inter-annotator agreement checks, reviewed edge cases, and conducted periodic audits, achieving over 95% labeling accuracy and minimizing bias in the dataset

In this project, I worked on training a natural language processing model to classify customer support queries, with the goal of enabling the AI to automatically categorize requests and suggest relevant responses. My role focused on annotating and labeling thousands of tickets, identifying intents and entities, normalizing text, and ensuring consistent application of labeling guidelines. The dataset included over 50,000 entries across multiple languages and product lines, and I personally annotated around 15,000, collaborating closely with a team of four to ensure coverage. To maintain high quality, we followed strict annotation protocols, performed inter-annotator agreement checks, reviewed edge cases, and conducted periodic audits, achieving over 95% labeling accuracy and minimizing bias in the dataset

2024

Education

Work History