M.S. student at NYCU · AI/ML Systems · Hyperparameter Optimization · Federated Learning · LLM Agents

I am a graduate student at National Yang Ming Chiao Tung University (NYCU), focusing on AI/ML systems, hyperparameter optimization, federated learning, and LLM-based research tooling.

My current work explores how LLM agents can assist Bayesian Optimization and Federated Learning Hyperparameter Optimization under limited evaluation budgets.

I enjoy turning research ideas into reproducible software systems, including experiment pipelines, visualization tools, and AI/ML prototypes.

• Hyperparameter Optimization (HPO)
• Bayesian Optimization (BO)
• Federated Learning (FL)
• LLM Agents and AI Research Tooling
• AI/ML Experiment Systems
• Backend and Full-stack Prototyping
• Reproducible Research Workflows

• [Relax][Frontend][TFLite] Add Conv3D support
• [Relax][Frontend] Add TFLite Frontend Support for CONV_3D_TRANSPOSE
• [Relax][TFLite] Add gather frontend expected IRModule tests

• [Feature Request] Support top/bottom tab split panes inside the same browser window

I contribute to open-source ML systems and developer tools, mainly around TVM Relax TFLite frontend and AI-assisted developer workflows.

Interactive dashboard for analyzing and replaying hyperparameter optimization experiments.

• Built for inspecting optimization behavior across tasks, methods, and random seeds
• Supports experiment browsing, metric visualization, and search process replay
• Tech stack: TypeScript, React, Vite, Flask

Repository: HyperOpt-viz-Studio

Deepfake detection project using CLIP and parameter-efficient fine-tuning.

• Focused on cross-manipulation generalization
• Evaluated model performance with AUC, accuracy, and F1-score
• Tech stack: Python, PyTorch, CLIP, PEFT

Repository: deepfake_detection

• Designing LLM-assisted Bayesian Optimization workflows
• Building experiments for Federated Learning Hyperparameter Optimization
• Studying how LLM agents can make optimization decisions under limited budgets
• Improving research tooling for experiment tracking and visualization

I am interested in software engineering and AI/ML engineering roles where I can combine:

• machine learning knowledge
• system implementation ability
• backend / full-stack prototyping
• research-to-product development

I am especially interested in roles related to AI platforms, ML systems, LLM applications, developer tools, and intelligent automation.

• Email: willie103006@gmail.com
• LinkedIn: https://www.linkedin.com/in/hsuweicheng/
• GitHub: https://github.com/weicheng-hsu