韩福宇

Key Skills

Software

No software listed

Top Subject Matter

E-commerce – Product Categorization & Customer Support

General Data Annotation & Quality Control

Top Data Types

Image

Video

Text

Top Task Types

Question Answering

Text Summarization

Fine-tuning

Freelancer Overview

I am proficient in annotation standards for text, images, speech, and multimodal data. I am skilled in core processes such as classification, bounding-box annotation, entity extraction, dialogue annotation, and quality control acceptance, and can quickly adapt to the data needs of different business areas such as large model fine-tuning, autonomous driving, and AIGC. At the same time, I am familiar with annotation platform operations, data cleaning, sample selection, annotation rule formulation, and batch quality inspection. I can independently complete the full workflow from organizing raw data, breaking down annotation tasks, coordinating team annotations, to performing data compliance and anonymization, and delivering training datasets. My core strengths lie in strong rule comprehension, sensitivity to detail error correction, and standardized process thinking. I excel at quickly grasping industry annotation standards, identifying annotation ambiguities and dirty data, and improving annotation efficiency and data accuracy through optimizing annotation processes and formulating standardized SOPs. I have experience in adapting AI training data, understand the data formats for large model fine-tuning, dialogue corpus organization, and sample balancing. I can not only perform basic annotation tasks effectively but also optimize data quality based on model training outcomes to avoid model fitting bias caused by annotation errors. I possess practical implementation capabilities and AI business logic understanding, allowing me to quickly take on the full range of responsibilities as an AI training engineer.

Education

Work History