about
Hey there good people! I'm Nathan, a graduate from KAIST with a major in Computer Science and double major in Business Technology Management, specializing in AI. My interests lie in the intersection of HCI and AI.
Apart from research, I like working on real-world projects, blending theoretical knowledge with practical applications. Check out my side projects to see what I am up to.
Welcome to my corner of the internet. I'm glad you're here!
experience
Jul — Present Full Stack AI Engineer · Darwin Intelligence
- •Built production RAG systems for enterprise knowledge management with high-accuracy document retrieval
- •Developed speech AI solutions including voice authentication systems
- •Architected multimodal AI platforms integrating Weaviate, Chroma, and LangChain for semantic search
- RAG
- Speech AI
- LLMs
Dec — Aug 2024 Software Engineering Intern · Calici
- •Redesigned company website for improved UX and navigation
- •Built chatbot for user support and website guidance
- •Migrated chatbot from third-party service to in-house solution, reducing costs
- Web Development
Jan — Jul 2023 Research Intern · KIXLAB KAIST
- •Engineered KUIZ, a modularized multiple-choice question generation platform using LLMs
- •Conducted user research including need-finding, ideation, and prototyping
- •Performed usability testing to validate platform design
- HCI
- User Research
- LLMs
Jan — Mar 2022 Research Intern · ICLAB KAIST
- •Created dashboard to visualize smartphone and smartband data from test users
- •Tracked user location, activities, and behavioral patterns
- •Identified and flagged anomalies in user data
- Data Visualization
- Data Analysis
Feb — Dec 2020 Student Researcher · NMAIL KAIST
- •Replicated research on adversarial attack training using pretrained GANs
- •Implemented pretraining models for photo manipulation applications
- •Experimented with adversarial attack strategies
- Machine Learning
- Deep Learning
- GANs
projects
Browser extension that helps users recognize echo chambers and media bias by highlighting emotionally-charged language and suggesting diverse perspectives on news articles.
- TypeScript
- Browser Extensions
- NLP
Interactive visualization tool for exploring machine learning algorithms in real-time. Features multiple clustering methods (K-Means, DBSCAN) and classification algorithms (KNN, Decision Trees) with live parameter tuning and visual feedback.
- JavaScript
- Canvas
- Machine Learning
- Visualization
LSTM-based deep learning model for sentiment classification of IMDB movie reviews. Achieves high accuracy through sequence modeling and word embeddings.
- Python
- TensorFlow
- Keras
- LSTM
- NLP
Full-stack pizza ordering platform with RESTful API backend and React frontend. Features real-time menu management, order tracking, and seamless user experience for online pizza delivery.
- Django REST Framework
- React
papers
2024 L@S'24 Workshop on Learnersourcing: Student-generated Content @ Scale
Hyoungwook Jin, Haesoo Kim, Nathan Mekuria Haile, Soyeong Min, Juho Kim
This paper presents a novel approach to learnersourcing multiple-choice questions by leveraging Large Language Model (LLM) interventions. The approach encourages modular, student-generated content at scale, facilitating more efficient and effective question creation in educational settings.
- Learnersourcing
- LLMs