Jason Zhuang

Key Skills

Software

Appen

Clickworker

Data Annotation Tech

Other

Top Subject Matter

No subject matter listed

Top Data Types

Text

Audio

Image

Top Label Types

Question Answering

Evaluation Rating

Entity Ner Classification

Audio Recording

Classification

Data Collection

Freelancer Overview

I am a native Mandarin and Cantonese speaker with fluent English and a strong foundation in business process optimization, data analysis, and supply chain management. Over recent years, I have specialized in AI data annotation, linguistic quality review, and prompt response classification for both Chinese Mandarin and Cantonese. My experience includes evaluating LLM outputs for fluency, accuracy, and contextual relevance, applying my analytical and linguistic skills to improve AI understanding and localization quality. I am comfortable working with complex evaluation guidelines, English instructions, and collaborative annotation tools. My cross-domain expertise—spanning e-commerce, manufacturing, and logistics—enables me to assess AI model outputs for logical clarity, domain accuracy, and practical usability. I have successfully completed multiple AI annotation and training projects on platforms such as dataannotation.tech, clickworker.com, and Neevo, ensuring high-quality and culturally accurate results.

Labeling Experience

Chinese Mandarin evaluate AI responses

Data Annotation TechTextClassificationQuestion Answering

This project involved evaluating AI-generated responses to Chinese Mandarin prompts across multiple domains, including general knowledge, business, and technical topics. My role focused on assessing the fluency, accuracy, contextual relevance, and cultural appropriateness of LLM outputs, with a particular emphasis on bilingual consistency and business logic alignment. The project also required familiarity with prompt engineering, response classification, and error pattern identification, contributing to the continuous improvement of AI model performance in Chinese-language contexts.

This project involved evaluating AI-generated responses to Chinese Mandarin prompts across multiple domains, including general knowledge, business, and technical topics. My role focused on assessing the fluency, accuracy, contextual relevance, and cultural appropriateness of LLM outputs, with a particular emphasis on bilingual consistency and business logic alignment. The project also required familiarity with prompt engineering, response classification, and error pattern identification, contributing to the continuous improvement of AI model performance in Chinese-language contexts.

2025 - 2025

Mandarin Audio Prompt Comparison for Quality Evaluation

OtherAudioEvaluation Rating

Compared Mandarin audio outputs from AI models to evaluate fluency, clarity, and contextual relevance. Applied detailed rubrics to assess pronunciation accuracy, tone naturalness, and semantic alignment with prompts. Provided structured ratings and feedback to improve model performance in Chinese speech generation. Comfortable working with bilingual guidelines and nuanced linguistic criteria.

Compared Mandarin audio outputs from AI models to evaluate fluency, clarity, and contextual relevance. Applied detailed rubrics to assess pronunciation accuracy, tone naturalness, and semantic alignment with prompts. Provided structured ratings and feedback to improve model performance in Chinese speech generation. Comfortable working with bilingual guidelines and nuanced linguistic criteria.

2025 - 2025

Picture collection project

OtherImageClassificationData Collection

Building, food and sign picture collection

Building, food and sign picture collection

2025 - 2025

Chinese Mandarin Audio Prompt Transcript Review

Data Annotation TechAudioEntity Ner ClassificationEvaluation Rating

This project involved transcribing and evaluating audio clips to support AI model training. I performed detailed transcription of spoken content, rated audio quality, and assessed speaker accents for clarity and consistency. Additionally, I reviewed background noise levels and inferred speaker environments (e.g., indoor, outdoor, public spaces) to help improve model robustness across varied acoustic conditions. The task required linguistic precision, acoustic awareness, and consistent application of evaluation standards. My contributions enhanced the accuracy of speech recognition systems and the diversity of training data for multilingual AI applications.

This project involved transcribing and evaluating audio clips to support AI model training. I performed detailed transcription of spoken content, rated audio quality, and assessed speaker accents for clarity and consistency. Additionally, I reviewed background noise levels and inferred speaker environments (e.g., indoor, outdoor, public spaces) to help improve model robustness across varied acoustic conditions. The task required linguistic precision, acoustic awareness, and consistent application of evaluation standards. My contributions enhanced the accuracy of speech recognition systems and the diversity of training data for multilingual AI applications.

2025 - 2025

Chinese Q&A Rating and Review

Data Annotation TechTextQuestion AnsweringEvaluation Rating

This project involved evaluating AI-generated answers to Chinese-language questions, often accompanied by contextual images. My role was to rate the quality, relevance, and accuracy of each response based on predefined criteria. Tasks included identifying whether the answer correctly addressed the question, assessing linguistic clarity, and determining alignment with visual cues when applicable. I applied consistent rating standards and provided feedback to support model refinement. The project emphasized attention to detail, cultural nuance in language understanding, and balanced judgment across diverse question types.

This project involved evaluating AI-generated answers to Chinese-language questions, often accompanied by contextual images. My role was to rate the quality, relevance, and accuracy of each response based on predefined criteria. Tasks included identifying whether the answer correctly addressed the question, assessing linguistic clarity, and determining alignment with visual cues when applicable. I applied consistent rating standards and provided feedback to support model refinement. The project emphasized attention to detail, cultural nuance in language understanding, and balanced judgment across diverse question types.

2025 - 2025

Education

Work History