Key Skills

Freelancer Overview

I have extensive experience in AI training data creation and data labeling through my research in AI4Science and computational materials modeling. I built and curated multiple large-scale datasets derived from DFT simulations, including atomic structures, magnetic moments, and energy relationships. Each dataset required precise data cleaning, feature labeling, and metadata annotation to enable effective machine learning model training and validation. My workflow involved defining labeling rules, ensuring inter-sample consistency, and integrating human-verified ground-truth data with automated scripts in Python (pandas, NumPy). What sets me apart is my ability to bridge domain-specific scientific expertise with ML data pipeline engineering. I designed data schemas for atomic environments, automated label generation for cation spin configurations, and validated data quality using cross-validation and model performance metrics (R², RMSE). These skills translate directly to AI data labeling and model training, where I can ensure accuracy, consistency, and interpretability of labeled datasets across complex technical domains.

Labeling Experience

Education

Work History