Welcome, my DGX Spark!
NLP & AI Research · 최재영
Jae Y. Choi
Computer Science undergraduate at DGIST
I research large language models, multimodal AI, and human-computer interaction. My work includes building training and evaluation pipelines, developing LLM agents, curating research data, and studying how people and research communities engage with AI systems.
News
Recent news
-
-
Joined the NVIDIA DLI Ambassador Meetup.
-
Submitted TimeDuet to ICASSP 2027.
-
–
Taught the NVIDIA DLI RAG course at KSA.
-
Joined CoRE-AI Lab.
-
Mentored MABC 2026 regional training participants.
-
Visited Lablup HQ.
Interests
Research and exploration
My work so far has spanned large language models, multimodal AI, and human–computer interaction. But my curiosity doesn’t stop there: I enjoy figuring out how computers and AI work, building something new, and taking on unfamiliar challenges.
Large Language Models
Built LLM agents and RAG systems, and worked on training pipelines and model evaluation.
Multimodal AI
Worked on audio-visual temporal grounding, dataset construction, and multimodal model training and evaluation.
Human–Computer Interaction
Explored interdisciplinary research practices and how researchers’ backgrounds shape their work.
Research
Submitted research
TimeDuet: A Benchmark for Audio-Visual Temporal Grounding on Misaligned Streams
* Equal contribution · DGIST & Hanyang University
Experience
Research and professional experience
Research Intern Current
CoRE-AI Lab, DGIST
Conducting research on AI models for predicting aroma molecules in mixtures.
DLI Certified Instructor Current
NVIDIA Deep Learning Institute
Certified instructor for the Building RAG Agents with LLMs course.
University Ambassador Current
NVIDIA Deep Learning Institute
Research Intern
LaMI Lab, DGIST
Conducted research on large language models and multimodal systems.
AI Ambassador
Upstage
Served as an AI developer-community ambassador and supported technical outreach.
Research Intern
SEMES · Smart EQ SW Center
Conducted R&D for an LLM agent service for semiconductor manufacturing equipment. Built multi-agent systems with LangChain and researched RAG-based domain knowledge methods.
Research Intern
DIAG · Yonsei University
Studied the interdisciplinary identities of HCI researchers and contributed to an ACM CHI manuscript in preparation.
Research Intern
DIAG · DGIST
Supported an AI-driven audit system, augmented linguistic corpus data for an NIA-funded project, and reviewed HCI methodology literature.
Project Manager, Simulation Platform
Quester
Initiated and led a Unity-based virtual experiment platform, coordinating developers, domain experts, vendors, deployment, and pilot trials.
Projects
Selected projects and leadership
DGIST AI Build Week 2026
Lead Organizer · Jul 2026Initiated and led a two-day, campus-wide AI festival with keynotes, hands-on workshops, sponsor booths, and an overnight team hackathon.
- Recruited and coordinated industry partners including Upstage, Lablup, and UNIVA.
- Designed the full program, including internship pathways at sponsor companies.
Play the LAB (PLAB)
Project ManagerLed end-to-end development of an interactive Unity platform for immersive virtual science experiments.
- Owned the project lifecycle from proposal and roadmap to deployment.
- Piloted with users at three sites and demonstrated at Ewha Womans University and Seoul Women's University.
NAVER CONNECT SW Camp
Team LeadLed a volunteer team through summer and winter camps across several schools, including program design and social-media outreach.
- The team was selected as an Outstanding Club by NAVER CONNECT.
Education
Education
Daegu Gyeongbuk Institute of Science & Technology
B.S. in Computer Science
GPA 3.79 · Six semesters completed · Daegu, South Korea
Kyungpook National University
Summer Credit Exchange
School of Computer Science and Engineering · GPA 3.50
Seoul National University
Winter Credit Exchange
Department of Electrical and Computer Engineering · GPA 3.85
Recognition
Certifications and awards
Certifications
Building RAG Agents with LLMs
NVIDIA DLI
Getting Started with AI on Jetson Nano
NVIDIA DLI
H-Mobility: Autonomous Driving — Perception
Hyundai NGV
Building Transformer-Based NLP Applications
NVIDIA DLI
Building Conversational AI Applications
NVIDIA DLI
Computer Vision for Industrial Inspection
NVIDIA DLI
Awards & Honors
Outstanding Training Award
Korea Air Force Basic Military Training Center
Outstanding Club Award
NAVER CONNECT · SW Sprout Camp
Hanwha Science Challenge — Bronze
Hanwha Group
Youth Content Dream Project — Grand Prize
Chungnam Content Korea Lab
Lectures
Teaching and lecture materials
– · 14:00–18:00 KST · Live online
Building RAG Agents with LLMs
Instructor · Korean Standards Association (KSA)
2026 ICT Innovation Square Expansion Program · NVIDIA DLI-based Industrial AI Transformation (AX) Challenge
- LLM inference interfaces and microservices.
- LLM pipelines with LangChain, Gradio, and LangServe.
- RAG agent implementation, modularization, and evaluation.
Building RAG Agents with LLMs
NVIDIA Deep Learning Institute · DLI Certified Instructor & University Ambassador
Access lecture materials and course resources on the KSA course website.