Joseph Smyth

Key Skills

Software

Data Annotation Tech

Top Subject Matter

No subject matter listed

Top Data Types

Text

Audio

Video

Top Label Types

Question Answering

Text Generation

Evaluation Rating

Emotion Recognition

Segmentation

Fine Tuning

Prompt Response Writing SFT

Freelancer Overview

I have hands-on experience as an AI Trainer and Rater, where I evaluated AI-generated responses, created rubrics for effective answers, and assessed the quality of other workers’ tasks. My background in Python and Java, along with my game programming education, has given me a strong technical foundation for data annotation and labeling work. I am a quick learner, dependable, and detail-oriented, which allows me to deliver high-quality results in fast-paced environments. My experience managing teams and working in customer service has also helped me develop strong communication and organizational skills, making me well-suited for roles involving AI training data and evaluation.

Labeling Experience

AI shopping Screenshot Evaluation

Data Annotation TechImageQuestion AnsweringText Generation

This tasks involved rating two AI responses to a prompt. We would look specifically at what the Response provided as recommendations. and rate the quality of the response. the we would fact check all of the products to make sure they were real and not just hallucinated by the model.

This tasks involved rating two AI responses to a prompt. We would look specifically at what the Response provided as recommendations. and rate the quality of the response. the we would fact check all of the products to make sure they were real and not just hallucinated by the model.

2025 - 2025

Generate Prompts and Rubrics for high Quality AI Responses

Data Annotation TechTextFine TuningEvaluation Rating

Write high quality rubric criteria to eliminate issues with AI responses to create ideal responses. This included multiple turns of user response and AI responses.

Write high quality rubric criteria to eliminate issues with AI responses to create ideal responses. This included multiple turns of user response and AI responses.

2025 - 2025

Providing Video and Segmenting Video

Data Annotation TechVideoSegmentation

For this project, the workers provided video from different length groups (30s-1:00, 1:00-2:00 3 min+). Then workers would then segment other workers videos, which could range from cooking videos, meme videos, or even lets plays of games.

For this project, the workers provided video from different length groups (30s-1:00, 1:00-2:00 3 min+). Then workers would then segment other workers videos, which could range from cooking videos, meme videos, or even lets plays of games.

2025 - 2025

AI Audio Response Evaluation

Data Annotation TechAudioQuestion AnsweringEmotion Recognition

This project involved Listening to a user audio prompt, the listening to AI Audio Responses. I looked for audio distortions or glitches, or if there were any incorrect information provided.

This project involved Listening to a user audio prompt, the listening to AI Audio Responses. I looked for audio distortions or glitches, or if there were any incorrect information provided.

2025 - 2025

AI Text response Evaluation

Data Annotation TechTextQuestion AnsweringText Generation

This project focused on the Truthfulness of AI Agents responses to a mix of real user prompts, and worker generated prompts. Other aspects this project had were Responses, tone, conciseness, and also overall helpfulness.

This project focused on the Truthfulness of AI Agents responses to a mix of real user prompts, and worker generated prompts. Other aspects this project had were Responses, tone, conciseness, and also overall helpfulness.

2025 - 2025

Education

Work History