Lab Profile · 2026
CCL.
GACHON UNIVERSITY
DEPARTMENT OF COMPUTER ENGINEERING

LABORATORY PROFILE / 2026

Cognitive
Computing
Lab.

Gachon University

Exploring federated learning, AI agents, healthcare AI and intelligent systems through AI and cognitive computing.

Federated learning across mobile, manufacturing and robotics, with local data and model exchange
AI, CONNECTED TO THE REAL WORLD.CCL

About the Lab

01 / OVERVIEW

Led by Professor KangYoon Lee in Gachon University's Department of Computer Engineering, CCL studies real-world challenges with AI, cognitive computing, IoT and cloud technologies.

Our FedOps platform connects model training and operations in distributed environments. We study post-TAVR complication and psychiatric risk prediction, alongside industry collaborations and national research projects.

Professor KangYoon Lee

PRINCIPAL INVESTIGATOR

KangYoon LeeProfessor

Department of Computer Engineering
Gachon University

keylee@gachon.ac.kr
Professor Profile ↗

Research Areas

02 / RESEARCH
01

Federated Learning

Collaborative training and lifecycle management in distributed environments. FedOps connects research with model operations.

02

AI Agent & LLM

Building, using and improving LLMs and Tool AI, with research on on-device models and efficient fine-tuning.

03

Healthcare AI

Medical language models, medical twins and risk prediction, including post-TAVR complications and mental health.

04

Intelligent Systems

IoT and cloud data applications, connecting sensor and multimodal data with intelligent systems.

COGNITIVE COMPUTING LAB · GACHON UNIVERSITY01 / 02
CCL.
RESEARCH PLATFORM
PUBLICATIONS & CONTACT

RESEARCH SOFTWARE

FedOps1.3

Federated AI AgentOps

From local data
to federated AI agents.

Train models together while keeping raw data on participants' devices. Use the trained models to build, run and improve AI agents.

  1. 01 / PREPARELocal setupConnect data
    Code · Model · Environment
  2. 02 / COLLABORATEFederated learningJoin a Task
    Aggregate model updates
  3. 03 / VERSIONModel versionsGlobal Model
    Version and lineage tracking
  4. 04 / BUILD & USEAgent useBuild · Test
    Local Chat · Serving

Web · Registry coordinates Tasks, participation and model versions. Agent Studio supports local development and execution. FL Server aggregates model updates.

Raw data stays local. Each Agent Build pins a selected model version. After validation, a new model can be used in a new Agent Revision.

Selected Publications

03 / PUBLICATIONS
2026

BitLoRA: Quantization-Compatible Adapter Tuning for 1.58-bit LLM in Federated On-Device AI-Agent

InSeo Song, KangYoon Lee · Expert Systems with Applications 311, 131397

2026

FedTFT: Federated Temporal Fusion Transformer for Interpretable Multi-Horizon Psychiatric Risk Prediction Across Cross-Silo Hospitals

Akeel Ahamed, Ri-Ra Kang, KangYoon Lee · IEEE Journal of Biomedical and Health Informatics

2024

FedOps: A Platform of Federated Learning Operations With Heterogeneity Management

Jihwan Moon, Semo Yang, KangYoon Lee · IEEE Access 12, 4301–4314

Contact CCL

04 / CONTACT

Professor KangYoon Lee

keylee@gachon.ac.kr

031-750-8902 · Office: AI Engineering Building, Room 312

Gachon University · Global Campus

CCL · AI Engineering Building, Room 103
1342 Seongnam-daero, Sujeong-gu, Seongnam-si, Gyeonggi-do 13120, Republic of Korea

Lab Website ↗

Sources: CCL Research · Publication · Professor · Contact / FedOps 1.3 Docs · Reviewed 2026.11.02

Download this HTML file or print / save it as a two-page A4 profile using the buttons above.