Drageon Lee

Drageon Lee

ML Researcher  ·  Computational Materials Informatics

Bridging battery materials domain expertise and data engineering with AI/ML research, with a focus on Computational Materials Informatics.

Google Scholar GitHub CV

Portfolio

Projects

Umicore  ·  2018-2021
⚗️

High-Nickel Cathode Materials R&D

  • Analyzed quality issues in high-nickel cathode materials: structural instability, oxygen loss, microcracks
  • Participated in tungsten additive development for high-nickel cathode optimization
  • Interpreted XRD and coin cell test results; authored 200+ quality & 8D reports
  • Analyzed lab-to-pilot scale-up gaps for process condition optimization
High-Ni Cathode XRD Six Sigma 8D Report
Battery Materials
Research
⚗️

Cathode Stability Prediction (CGCNN)

Graph-based ML for thermodynamic stability prediction of cathode materials using Materials Project and OQMD. Analyzed cross-database generalization and domain shift via zero-shot and transfer learning approaches.

CGCNN GNN Materials DFT
Materials AI
Research
🔋

Battery Materials Modeling: DFT + ML

DFT+U workflow with Quantum ESPRESSO for layered oxide cathode stability. Linked computational results to ML screening pipelines, exploring synthesis and cycling behavior through computational modeling.

DFT Materials Quantum ESPRESSO
Computational Chemistry
UT Austin
🧠

Post-stroke Prediction with GNN + Explainability

Graph Transformer (TransformerConv) model for post-stroke outcome prediction using ICD codes and clinical features. Benchmarked GNNExplainer, PGExplainer, GSAT, and GIB for clinical decision support.

GNN XAI Healthcare AI PyTorch
Clinical AI
UT Austin · Capstone
🔍

Hybrid Retrieval for Domain-Specific Search

End-to-end RAG system on AWS combining BM25 keyword search with vector search. Built with Terraform, OpenSearch Serverless, Lambda, Step Functions, and Bedrock. Agentic workflow for orchestration and evaluation.

RAG LangChain Bedrock AWS
Oxy Capstone GitHub ↗
UT Austin
🏥

Surgical Tool Search Agent

Search agent for surgical and medical tool queries using LangChain agents. Designed retrieval pipeline for accuracy, usability, and workflow efficiency in medical information retrieval.

LangChain Agents RAG
Healthcare AI
UT Austin
🛠️

LLM-based QA Tool & Defect Analytics

LLM-powered QA automation for error detection, root cause classification, and triage support. Includes Power BI dashboard for defect analytics and operational priority visualization.

LLM Power BI Analytics
QA Automation
UT Austin
📚

Guided Learning Evaluation / LLM Judge

Evaluated Gemini's guided tutoring mode against Bloom's Taxonomy across 60 tasks. Designed LLM Judge agent assessing pedagogical alignment, Socratic guidance, and conceptual reinforcement for SQL learning.

Gemini LLM Judge Evaluation
EdTech AI
UT Austin
🎵

Multimodal Music Emotion Recognition

Audio-text multimodal model for music emotion recognition using AST and BERT. Evaluated on MERGE and PMEmo datasets, extended to cross-modal retrieval and emotion-based music recommendation via cosine similarity.

AST BERT Multimodal
Multimodal AI
Circle Platform
⚡

API Platform & Demand Forecasting Engine

Developed 100+ APIs accelerating data collection 3x. Built AWS data pipelines, MariaDB query optimization (50% faster processing), and demand forecasting models with Scikit-learn that grew the user base from 1,100 to 2,300.

AWS MariaDB Scikit-learn Pipeline
Data Engineering
ABPA
📊

ABPA Connect — Data Integration Pipeline

Unified Givebutter and EveryAction data into BigQuery as a single source of truth. Cleaned CRM/fundraising data, resolved cross-platform duplicates, and built Looker Studio KPI dashboards for end-to-end reporting.

BigQuery Looker Studio ETL
Data Engineering
Circle Platform
🖼️

Image AI — Diffusion & Style Feature Dev

Evaluated and improved AI image editing quality including object/background removal, super-resolution, and diffusion-style editing. Automated manual QA processes and iterated through tuning/testing for stable output.

Diffusion OpenCV Image QA
Computer Vision

Career

Experience

August 2024 – Present

Graduate Researcher & AI Lab Instructor

The University of Texas at Austin

  • Led AI lab sessions on Hugging Face Transformers, LLM fine-tuning, computer vision, and explainable AI
  • Mentored student teams on EHR analysis, medical NLP, and healthcare AI projects
  • Developed coursework and materials for SQL, GenAI, and model evaluation
  • Conducted research on GNN-based clinical prediction, multimodal AI, and RAG systems
  • Built and evaluated LLM-based tools for QA automation and tutoring assessment

April 2022 – June 2024

Data Part Leader

Circle Platform

  • Developed 100+ APIs accelerating data collection speed by 3× (10 min → 3 min)
  • Built AWS-based data pipelines and improved MariaDB event log performance by 50%
  • Designed demand forecasting models with Scikit-learn, driving user base growth from 1,100 to 2,300
  • Evaluated and improved AI image editing features: object removal, super-resolution, diffusion styling
  • Implemented pseudonymization techniques using Hadoop MapReduce for large-scale data
  • Conducted consumer behavior sequence analysis and optimized search/pattern algorithms

August 2018 – October 2021

Battery Materials Domain Expert  · Quality Engineer

Umicore

  • Analyzed quality issues in high-nickel cathode materials: structural instability, oxygen loss, microcracks
  • Characterized moisture effects, residual lithium imbalance, and Li₂CO₃/LiOH-related defects using XRD and coin cell tests
  • Participated in tungsten additive development for high-nickel cathode optimization
  • Analyzed lab-to-pilot scale-up gaps and process condition optimization
  • Authored 200+ quality reports including 8D analysis; led Japan customer audits
  • Applied Six Sigma for defect rate management and process control
🔗 This hands-on expertise in battery cathode materials directly informed the GNN-based Cathode Stability Prediction and DFT+ML research projects, bridging real-world materials knowledge with computational AI.

Academic

Education

M.S. in Information Studies

The University of Texas at Austin

August 2024 – Expected May 2026

Data Management · Machine Learning · Natural Language Processing · Cloud Infrastructure

B.S. in Fine Chemical Engineering

Chungnam National University

March 2010 – August 2017

Mathematics · Statistics · Chemical Engineering · Biological Engineering

Technical

Skills

Languages

Python R SQL JavaScript MATLAB

ML / DL

PyTorch TensorFlow Scikit-learn PyTorch Geometric Hugging Face XGBoost NLTK OpenCV

LLM & GenAI

LangChain OpenAI API Gemini AWS Bedrock RAG Prompt Engineering LLM Evaluation

Data Engineering

pandas NumPy MySQL MongoDB BigQuery MariaDB Looker Studio Power BI Hadoop

Cloud & Infrastructure

AWS GCP Docker Terraform AWS Lambda OpenSearch Step Functions

Materials & Scientific

Quantum ESPRESSO DFT+U XRD Analysis CGCNN Six Sigma 8D Report