Taehee Kim

Key Skills

Software

Appen

Clickworker

CrowdSource

Data Annotation Tech

Google Cloud Vertex AI

Labelbox

Lionbridge

OneForma

Remotasks

Scale AI

Telus

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Video

Top Task Types

Audio Recording

Evaluation/Rating

Prompt + Response Writing (SFT)

Text Generation

Translation/Localization

Freelancer Overview

have extensive experience in AI training data creation, linguistic annotation, and evaluation, specializing in multilingual (Korean–English–Spanish) datasets for NLP, machine translation, and generative AI models. My work includes building structured Excel-based labeling frameworks, creating rubrics for model evaluation, and performing semantic, fluency, and factuality QA across engines. I’ve led large-scale projects involving translation quality assessment, intent classification, and response ranking, applying both linguistic expertise and analytical precision to improve model output accuracy and consistency. What sets me apart is my ability to bridge linguistic, cultural, and technical insight — combining over a decade of professional translation and localization experience with advanced understanding of AI prompt design, model evaluation logic, and automation workflows. I bring a data-driven mindset, meticulous attention to detail, and strong familiarity with annotation guidelines, GenAI output testing, and rubric creation for multilingual systems, ensuring that every dataset I deliver contributes directly to measurable model improvement.

Labeling Experience

Multilingual AI Data Labeling & Evaluation Project

Scale AITextClassificationQuestion Answering

Contributed to large-scale multilingual AI training datasets for NLP and generative AI evaluation. Performed fine-grained text classification, intent tagging, and translation quality assessment (accuracy, fluency, adequacy) for Korean–English–Spanish corpora. Designed evaluation rubrics, scored model outputs, and verified data consistency across engines. Used structured Excel templates to log and validate annotations, ensuring inter-annotator agreement and guideline compliance. Supported model alignment through prompt-response scoring, contextual judgment, and qualitative feedback summaries.

Contributed to large-scale multilingual AI training datasets for NLP and generative AI evaluation. Performed fine-grained text classification, intent tagging, and translation quality assessment (accuracy, fluency, adequacy) for Korean–English–Spanish corpora. Designed evaluation rubrics, scored model outputs, and verified data consistency across engines. Used structured Excel templates to log and validate annotations, ensuring inter-annotator agreement and guideline compliance. Supported model alignment through prompt-response scoring, contextual judgment, and qualitative feedback summaries.

2023 - 2025

Education

Work History