Building clinically grounded decision-support systems for safer medication use in real-world care.

I work on medication safety, polypharmacy, and decision support at the intersection of pharmacy, regulatory strategy, and analytics.

My goal is not just to build models, but to build systems that clinicians can actually trust, interpret, and use in practice — because in healthcare, “probably works” is not a design standard.

• Medication safety and safer medication use
• Polypharmacy and complex medication decisions
• Guideline-grounded and knowledge-guided decision support
• Patient-specific modeling for treatment and dosing
• Healthcare systems that are usable, traceable, and safety-aware

An interactive medication safety prototype built on public Korean DUR contraindication data, designed to make drug–drug risk relationships more visible and easier to explore. The project also highlights vulnerable population signals to support more intuitive safety-oriented reasoning.

Tech: Python · Streamlit · NetworkX · Plotly · Medication safety data

A clinically oriented reasoning system integrating drug knowledge and guideline information to support more structured and traceable treatment decisions.

Tech: Python · AWS · Knowledge Graph · RAG · LLM · React

A healthcare data reproduction and evaluation project focused on benchmarking, downstream utility, and practical model assessment for synthetic clinical data.

Tech: Python · Evaluation Pipelines · Healthcare Data · Reproducibility

Languages
Python · R · SQL · JavaScript

Modeling & Analysis
scikit-learn · PyTorch · statistical analysis · simulation modeling · clustering

Healthcare AI & Systems
RAG · Knowledge Graphs · clinical reasoning workflows · patient-specific modeling

Infrastructure & Apps
AWS · React · serverless architecture

• M.S. in Analytics — Georgia Institute of Technology
• Licensed Pharmacist (Korea)
• Former Regulatory Affairs Specialist in pharmaceutical development

That background shapes how I approach healthcare AI:
I care less about impressive models in isolation, and more about clinically meaningful systems that can hold up in real decision-making environments.

I’m especially interested in building systems for:

• safer medication decisions
• high-risk and complex patients
• interpretable clinical decision support
• real-world healthcare implementation

Long term, I want to help build healthcare systems that are not only technically strong, but also safe, practical, and worth trusting.