Skills

Overview

I build data-driven systems that are reliable, reproducible, and practical for real-world use. My work combines machine learning research, engineering execution, and stakeholder-facing analytics.

Core Research and Modeling

• Benchmarking and leaderboard design for ML systems
• Model evaluation, error analysis, and reliability assessment
• Interpretable machine learning for decision support
• Multi-fidelity and multimodal modelling workflows
• LLM application workflows for practical tasks
• GNN-based modelling for structured and materials-related data

Engineering and Infrastructure

• Python (PyTorch, TensorFlow, scikit-learn)
• GPU training with CUDA
• HPC workloads and job orchestration with SLURM
• YAML-configurable training pipelines for reproducible experiments
• Linux and Bash scripting automation
• Git-based collaborative development workflows

Data and BI Delivery

• SQL and Python for production-oriented analytics workflows
• Dashboarding and reporting with QuickSight, Tableau, and PowerBI
• Cross-functional communication with technical and business stakeholders

Languages

• English (professional proficiency)
• Chinese (native)