Key Skills

Software

CrowdSource

Other

Internal/Proprietary Tooling

Scale AI

Top Subject Matter

Cloud Data Analysis for AI Models

Metadata Analysis for Online Content Moderation

AI Training in E-commerce and Digital Advertising

Top Data Types

Image

Text

Video

Top Task Types

Bounding Box

Classification

Diagnosis

Evaluation Rating

Prompt Response Writing SFT

Freelancer Overview

I am an experienced professional with a strong background in data analysis, cloud computing, and AI training. Over the past several years, I've honed my skills in metadata analysis and data labeling, particularly through my work on Google’s Trust & Safety project, where I examined and annotated metadata for video audits. My role involved detailed data analysis and collaboration with cross-functional teams to ensure accuracy and compliance, which has equipped me with the precision and attention to detail necessary for high-quality data labeling tasks. Currently, I am engaged in remote AI training for Korean writers, where I contribute to various projects aimed at enhancing generative AI models. My work involves ranking AI-generated responses, creating original content based on given topics, and evaluating the factual accuracy of AI-produced text. This role has allowed me to apply my language skills and creative writing abilities to improve the performance of AI models, ensuring that they generate more accurate and contextually appropriate content. My experience in this field, combined with my strong background in data analysis, enables me to bring a unique perspective to AI training projects, particularly in the context of language processing and content generation.

Labeling Experience

AI Text Generation and Evaluation for Korean Language Models

Scale AITextBounding BoxText Generation

This project involved working on various Korean language AI models with the objective of improving their ability to generate contextually accurate and coherent text. My role included generating prompts and responses, evaluating AI-produced text for accuracy and relevance, and classifying text based on specific criteria. The tasks were focused on enhancing the AI's understanding of the Korean language, cultural nuances, and content appropriateness. I worked with large datasets, ensuring high-quality annotations by adhering to strict guidelines and quality measures. The project was ongoing and required continuous collaboration with the AI development team to refine the models based on the latest evaluations.

This project involved working on various Korean language AI models with the objective of improving their ability to generate contextually accurate and coherent text. My role included generating prompts and responses, evaluating AI-produced text for accuracy and relevance, and classifying text based on specific criteria. The tasks were focused on enhancing the AI's understanding of the Korean language, cultural nuances, and content appropriateness. I worked with large datasets, ensuring high-quality annotations by adhering to strict guidelines and quality measures. The project was ongoing and required continuous collaboration with the AI development team to refine the models based on the latest evaluations.

2023

Metadata Analysis and Content Moderation for Google Trust & Safety

Internal Proprietary ToolingVideoClassificationText Summarization

I was part of the Google Trust & Safety team, where I was responsible for analyzing metadata associated with video content to ensure compliance with Google's policies. The project involved meticulous review and annotation of video data to identify and classify content according to safety guidelines. I collaborated closely with the Google Ads team to provide feedback on policy enforcement and contribute to the development of safer online environments. The project was extensive, involving the classification of thousands of videos and the evaluation of content for potential policy violations. My role required strong attention to detail, proficiency in data analysis, and a deep understanding of content moderation principles.

I was part of the Google Trust & Safety team, where I was responsible for analyzing metadata associated with video content to ensure compliance with Google's policies. The project involved meticulous review and annotation of video data to identify and classify content according to safety guidelines. I collaborated closely with the Google Ads team to provide feedback on policy enforcement and contribute to the development of safer online environments. The project was extensive, involving the classification of thousands of videos and the evaluation of content for potential policy violations. My role required strong attention to detail, proficiency in data analysis, and a deep understanding of content moderation principles.

2019 - 2022

Education

Work History