Evan Park

Key Skills

Software

Appen

Labelbox

Prodigy

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Document

Geospatial Tiled Imagery

Text

Top Task Types

Prompt Response Writing SFT

Text Generation

Text Summarization

Translation Localization

Freelancer Overview

As a bilingual data annotator fluent in English and Korean, I have hands-on experience contributing to high-impact AI training projects. My core strengths include evaluating AI-generated text (LLM outputs), labeling user intent, and sentiment analysis in both languages. I’ve supported training pipelines for models similar to GPT by ranking and comparing responses with high consistency and contextual understanding. My background in linguistics and natural language processing gives me an edge when working with diverse data. I am familiar with tools such as Labelbox, Prodigy, and Scale ai, and I’m comfortable following detailed guidelines to ensure high annotation quality and consistency across large datasets.

Labeling Experience

LLM response Evaluation for bilingual AI Assistant

Scale AITextClassificationEmotion Recognition

Contributed to the development of an English-Korean bilingual language model by evaluating AI-generated responses based on helpfulness, relevance, tone, and accuracy. Tasks included ranking model outputs, rewriting user prompts to increase clarity, and identifying emotional tone in text. Worked with over 15,000 prompt-response pairs while maintaining over 98% QA score. Ensured adherence to detailed task guidelines and adjusted labeling strategy to accommodate model updates. Work contributed directly to fine-tuning conversational AI models.

Contributed to the development of an English-Korean bilingual language model by evaluating AI-generated responses based on helpfulness, relevance, tone, and accuracy. Tasks included ranking model outputs, rewriting user prompts to increase clarity, and identifying emotional tone in text. Worked with over 15,000 prompt-response pairs while maintaining over 98% QA score. Ensured adherence to detailed task guidelines and adjusted labeling strategy to accommodate model updates. Work contributed directly to fine-tuning conversational AI models.

2024 - 2024

Education

Work History