Shirley Huang

Key Skills

Software

Data Annotation Tech

Label Studio

OneForma

SuperAnnotate

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Audio Recording

Data Collection

Evaluation Rating

Prompt Response Writing SFT

Translation Localization

Freelancer Overview

LLM Evaluation and Text Generation Specialist skilled in training generative AI models across Cantonese, Mandarin, and English. Adept at translation and localization with a strong focus on cross-linguistic accuracy and cultural relevance. Demonstrated ability to manage AI training projects end-to-end, delivering high-quality outcomes on time while strengthening collaborative and client relationships.

Labeling Experience

Voice AI Trainer

Internal Proprietary ToolingAudioPrompt Response Writing SFTAudio Recording

•Contributed over 1500 minutes of high-quality natural dialogue audio to enhance multilingual AI model training and inclusivity. •Operated professional recording equipment in soundproof environments to deliver clear and compliant audio materials. •Collaborated within a multilingual team to align recording standards and improve dataset consistency through feedback. •Followed project guidelines strictly to ensure data accuracy and diversity. •Provided authentic speech data covering daily and cultural topics to support voice recognition technology.

•Contributed over 1500 minutes of high-quality natural dialogue audio to enhance multilingual AI model training and inclusivity. •Operated professional recording equipment in soundproof environments to deliver clear and compliant audio materials. •Collaborated within a multilingual team to align recording standards and improve dataset consistency through feedback. •Followed project guidelines strictly to ensure data accuracy and diversity. •Provided authentic speech data covering daily and cultural topics to support voice recognition technology.

2025

AI Multilingual Generalist

Data Annotation TechTextTranslation LocalizationEvaluation Rating

Performed multilingual (Cantonese/Mandarin/English) data annotation and model evaluation for AI training platforms for over a year, enhancing model accuracy and cultural relevance across diverse language tasks. Assessed and rated AI-generated responses based on criteria including instruction following, truthfulness, verbosity, localization, writing quality, and harmfulness, while providing detailed feedback to improve model performance. Developed and tested challenging prompts—including sensitive or toxic topics—to identify model vulnerabilities, refine outputs, and strengthen response safety and appropriateness. Conducted fact-checking and research to verify response authenticity, and created high-quality dialogue datasets to train multilingual chatbot conversational abilities.

Performed multilingual (Cantonese/Mandarin/English) data annotation and model evaluation for AI training platforms for over a year, enhancing model accuracy and cultural relevance across diverse language tasks. Assessed and rated AI-generated responses based on criteria including instruction following, truthfulness, verbosity, localization, writing quality, and harmfulness, while providing detailed feedback to improve model performance. Developed and tested challenging prompts—including sensitive or toxic topics—to identify model vulnerabilities, refine outputs, and strengthen response safety and appropriateness. Conducted fact-checking and research to verify response authenticity, and created high-quality dialogue datasets to train multilingual chatbot conversational abilities.

2024

Education

Work History