刘哲

Key Skills

Software

Other

Label Studio

Top Subject Matter

LLM Output Evaluation

Chinese-English Long-form Text

Dialogue Domain Expertise

Top Data Types

Text

Computer Code Programming

Image

Top Task Types

Classification

Data Collection

Segmentation

Question Answering

Freelancer Overview

I am a Computer Science master’s student with hands-on experience in AI data processing, model evaluation, and training data preparation. I have worked on data cleaning, feature extraction, and structured dataset construction for machine learning tasks, with strong proficiency in Python (Pandas, NumPy) and data handling workflows. I also have experience in AI model evaluation and annotation-related tasks, including LLM output assessment, data labeling, and quality checking for structured and unstructured datasets. My background in machine learning and AI security research enables me to accurately follow annotation guidelines and deliver high-quality training data with attention to detail and consistency.

Labeling Experience

Algorithm Engineering Intern – Data Labeling & Training Data Support

TextClassification

Managed pre-training data cleaning and label construction for AI time-series prediction development in a smart hardware context. Processed and structured training samples for model input, performing denoising and anomaly detection using Python tools. Designed and applied classification schemes to enable automated result filtering for algorithm model evaluation. • Built data augmentation solutions including noise injection and sequence concatenation. • Assessed and validated training data quality to improve model accuracy and stability. • Processed structured time-series data for input to prediction models. • Supported experiment scaling to 10,000+ records through sampling and augmentation.

2026 - Present

LLM Response Evaluation & AI Content Assessment (Practice Project)

OtherText

Evaluated large language model outputs for quality, accuracy, and instruction adherence in diverse Chinese and English tasks. Performed response quality assessment, output ranking, and error classification on over 500 samples. Applied annotation guidelines for standardized labeling and rigorous quality checking. • Conducted content safety reviews, identified hallucinations, and flagged anomalies for further analysis. • Worked bilinguially with Chinese and English data for both long-form and multi-turn dialogue. • Performed preference comparison across task categories and classified errors systematically. • Supported RLHF fundamentals and contributed to prompt evaluation workflows.

2025 - Present

AI Training Data Processing & Quality Assurance (Practice Project)

OtherTextClassification

Performed large-scale text data cleaning, labeling schema design, and structuring for AI model training pipelines. Standardized and processed CSV/JSON datasets through deduplication, noise filtering, and feature extraction for classification tasks. Conducted quality assurance and data screening to optimize training data integrity and readiness for fine-tuning. • Built and organized labeled training samples for text classification. • Ensured rigorous pre-training data review and quality assurance protocols. • Used Python (Pandas/NumPy) for automated processing and data conversion. • Handled multi-source data conformity and enhanced input consistency.

2024 - 2025

Education

Work History