三张

Key Skills

Software

Labelbox

CVAT

Appen

Remotasks

Toloka

Top Subject Matter

LLM Evaluation

General NLP

Computer Vision

Top Data Types

Text

Image

Document

Top Task Types

RLHF

Bounding Box

Classification

Freelancer Overview

AI Data Annotator & LLM Evaluator. Core strengths include Labelbox, CVAT, and Appen. Education includes Bachelor of Arts, Beijing Language and Culture University (2023). AI-training focus includes data types such as Text and Image and labeling workflows including RLHF, Bounding Box, and Classification.

Labeling Experience

AI Data Annotator & LLM Evaluator

LabelboxTextRLHF

As an AI Data Annotator & LLM Evaluator at OpenTrain AI, I evaluated and ranked LLM-generated responses across eight quality dimensions. I authored high-quality prompts and responses for instruction tuning and flagged unsafe or biased content to improve model alignment. I consistently exceeded project quality thresholds with RLHF comparison tasks and contributed to safety and alignment improvements. • Evaluated 3,500+ RLHF comparison tasks with a 97.8% inter-annotator agreement rate. • Authored 500+ prompts and gold responses for LLM fine-tuning. • Identified harmful, biased, or policy-violating content for major AI clients. • Used Labelbox, Label Studio, and Amazon SageMaker Ground Truth for annotation.

2025 - Present

RLHF Preference Annotator

RemotasksTextRLHF

As an RLHF Preference Annotator with Outlier AI via Remotasks, I rated and compared dual LLM responses regarding quality and instruction-following. I wrote detailed justifications for preference choices to train reward models. I participated in specialized projects such as mathematics reasoning, coding review, and creative writing assessment. • Compared and rated multi-turn conversational LLM outputs for RLHF. • Provided justifications to enhance reward model optimization. • Contributed to domain-specific tasks in math, coding, and creative writing. • Used Remotasks and standard annotation tools for project delivery.

2024 - Present

Content Moderator & Quality Analyst (AI Training)

AppenTextClassification

As a Content Moderator & Quality Analyst with Appen, I reviewed and categorized AI-generated text content. I applied content policy frameworks and flagged violations for a major social media client. I provided structured feedback and contributed to policy clarifications. • Reviewed 8,000+ AI-generated content items for policy adherence. • Flagged misinformation, NSFW, and hate speech in LLM outputs. • Provided feedback leading to three policy clarifications. • Ensured 96%+ accuracy across daily moderation tasks with Appen tools.

2024 - 2024

Data Labeling Specialist

CVATImageBounding Box

In my role as a Data Labeling Specialist at DataForce by TransPerfect, I performed precise image and video annotation for computer vision training. I also annotated text samples for sentiment analysis, intent classification, and NER in English and Chinese. Additionally, I conducted quality audits on peer annotations and maintained high accuracy. • Annotated image and video data using bounding boxes, semantic segmentation, and keypoint labeling. • Completed 10,000+ text sample annotations for sentiment and NER tasks. • Conducted 200+ quality assurance audits, correcting discrepancies. • Leveraged CVAT and Labelbox for annotation tasks.

2024 - 2024

AI Content Annotator (Bilingual EN/ZH)

TolokaText

As an AI Content Annotator for Toloka AI, I evaluated English-Chinese machine translation outputs for fluency and adequacy. I created bilingual parallel corpora by translating and aligning sentence pairs. I also contributed to translation assessment for an NMT model improvement project. • Evaluated translation quality and cultural appropriateness of NMT outputs. • Aligned and translated over 2,000 sentence pairs for parallel corpus creation. • Ranked among the top 10% of annotators based on quality consistency. • Contributed insights for model improvement using Toloka platform.

2023 - 2024

Education

Work History