Suhyeon Jun 전수현

M.S. Student · Computer Science and Engineering

POSTECH · Machine Learning Lab (MLLab)

About

I am an M.S. student in the Department of Computer Science and Engineering at Pohang University of Science and Technology (POSTECH). I am a member of the Machine Learning Lab, where I am advised by Prof. Dongwoo Kim.

My research focuses on Trustworthy AI, with particular interest in methods for machine unlearning, privacy-preserving ML, and robust evaluation of AI systems. I am also broadly interested in Generative AI.

Timeline

2026
- Feb
  
  M.S. in Computer Science and Engineering current
  
  POSTECH
  
  Machine Learning Lab, advised by Prof. Dongwoo Kim
2025
- Feb – Jul
  
  AI Engineer
  
  Dalpha (Sprint Team)
  
  Building real-world AI services across LLMs, generative models, and multimodal applications.
2024
- Jul – 2025 Jan
  
  AI Engineer (Intern)
  
  Dalpha (Sprint Team)
- Aug
  
  B.S. in Smart ICT Convergence
  
  Konkuk University
  
  Minor in Biomedical Science and Engineering
2023
- Oct – Dec
  
  Research Intern
  
  POSTECH Institute of Artificial Intelligence (AIoT Lab)
  
  Text-prompted timbre generation using Normalizing Flow.
2022
- Jan – Dec
  
  Research Assistant / Intern
  
  KITECH, Digital Healthcare R&D Department
  
  Deep learning for printed spot color prediction.

Experience

Work

AI Engineer · Sprint Team

Dalpha · Seoul, South Korea

Jul 2024 – Jul 2025

Worked on AI projects across various domains, developing real-world AI services including generative AI (text/image, diffusion), TTS, image matching, chatbot, text search, information extraction, and survey automation.
- Built a Korean audiobook TTS system using 2,000+ hours of collected audio, with a 5-step preprocessing and 2-stage training pipeline for improved pronunciation and speaker style.
- Implemented a visual and textual similarity-based product matching pipeline for large-scale retrieval across 7.4M+ items.
- Developed a resume parsing solution that extracts structured entity information from unstructured documents.
- Built a survey automation system by generating a dynamic decision tree from PDF forms using OCR, question structuring, and semantic analysis of answer choices.
- Built a KakaoTalk-integrated chatbot with retrieval and LLM-based response generation.
Research Intern · AIoT Lab

POSTECH Institute of Artificial Intelligence · Pohang, South Korea

Oct – Dec 2023

Researched voice generation and conversion using prompt-conditioned timbre representation with Normalizing Flow.
- Converted an image-text cross-modal alignment structure (LNFMM) into a text-to-timbre mapping framework.
- Designed guidelines and constructed a prompt-driven dataset for timbre description and model training.
- Built a preprocessing and training pipeline to learn non-linear mappings between text and timbre embeddings using Normalizing Flow.
Contract Research Assistant / Research Intern · Digital Healthcare R&D Department

Korea Institute of Industrial Technology (KITECH) · Cheonan, South Korea

Jan – Dec 2022

Developed a deep learning model for spot color prediction in printing, achieving 154% higher accuracy over prior work and leading to a first-author publication.

Suhyeon Jun 전수현

About

Timeline

Experience

Work

Skills

Certificates

Languages