AN TAESEUNG

Andong National University

Bachelor’s degree in Electronic Engineering Education · GPA 3.49 / 4.5

• Graduation project: deep learning for apple disease detection.

RAG-Based Question Answering for Board Game Rules

Internal PoC during employment at Red Button

• Built and tested a LangChain pipeline for rulebook indexing, passage retrieval, and grounded answer generation.
• Observed limitations in interpreting user intent and multi-constraint queries; deferred rollout because the prototype did not achieve sufficient usability or answer quality.

Deep Learning for Apple Disease Detection

Graduation project · IITP / Korean Foundation for Quality AI training

• Handled data, modeling, and backend work: disease-image collection and validation, data pipelines, and CNN training experiments.
• Organized small models hierarchically; implemented quantization, TensorFlow Lite deployment, and an inference API for resource-constrained environments.

Structured-Document OCR & Data Pipeline

EPIKAR · Python, OpenCV, Tesseract, Flask

• Sole developer across collection, labeling, image generation, training, inference, APIs, and on-premises deployment; approximately 1,000 source and 100,000+ generated images.
• Reached approximately 93% structured-table recognition accuracy on an internal test set (target: 99%), using ground-truth scoring; assessed feasibility and cost against commercial OCR.

AI-Assisted Text Simplification for Slow Learners

Frontend development · Tech for Impact / Peach Market

• Built Peachseoga AI’s conversational React UI, administration and account screens, and API integration, considering slow learners’ cognitive characteristics; scope: frontend and service integration.

Red Button

Software Engineer · Digital Transformation Team

• Developed membership, coupon, reservation, and administration systems; operated AWS/Docker infrastructure, traced client–API issues, and built a RAG PoC.

Metapet

Researcher · Research & Development Team

• Built and deployed FastAPI/Vue/AWS MVPs, 3D/mobile AR prototypes with Three.js/WebRTC, and an O2O service. Employment across corporate transition: May–July 2022 / Aug 2022–Jan 2023.

EPIKAR Inc.

Software Engineer · Service Development Team

• Designed and operated OCR APIs and training-data pipelines; developed Vue.js frontend interfaces.

Mentor, K-Digital Training — Big Data Analytics

Korean Foundation for Quality

• Mentored learners in Python-based data warehousing and visualization.

Mentor, K-Digital Training — Full-Stack Development

Andong National University

• Mentored learners in Flask/Vue development and mini-projects.

Teaching Assistant & Laboratory Safety Supervisor

Gifted Education Center, Andong National University

• Activities included building a rotating 3D display with Arduino and Processing.

Industry-Specific AI Developer Training · IITP / Korean Foundation for Quality · 960 hours completed

Three university-level encouragement awards in capstone design, startup ideas, and startup achievement presentations