01
AI Systems & RAG
Efficient on-premises AI inference, retrieval-augmented generation, adaptive retrieval, reranking, and system-level optimization for generative models.
System Software Lab. studies system software for AI inference, machine-centric video, edge computing, storage, RAG, and resource optimization. We combine learning-based decision making with practical system design.
Efficient on-premises AI inference, retrieval-augmented generation, adaptive retrieval, reranking, and system-level optimization for generative models.
QoA-aware streaming, video analytics, edge/cloud collaboration, and multi-node orchestration designed for machine perception rather than only human viewing.
Video and vector storage, hot/cold data placement, adaptive replication, edge caching, SSD-aware allocation, and data placement under capacity constraints.
DRL and optimization techniques for compute, storage, bandwidth, power, and latency management across heterogeneous edge and cloud systems.
This project develops machine-centric video streaming and multi-node orchestration techniques that jointly optimize representation quality, network transfer, and compute placement for downstream AI inference. QoA-aware control adapts video quality and resource allocation to preserve analysis accuracy while reducing end-to-end latency, bandwidth, and computation overhead.
This project advances secure and efficient on-premises generative AI for financial environments while training researchers who can bridge AI models and system software. System Software Lab. focuses on system-level topics including efficient inference, RAG pipelines, resource management, storage and retrieval optimization, and practical deployment of generative models in constrained on-premises environments.
Department of Computer Engineering · Department of Artificial Intelligence Engineering
Research: system software, AI systems, multimedia/edge systems, storage, resource and energy optimization.
Students can participate in active research projects in AI systems, RAG, edge AI, machine-centric video, storage, and resource optimization. Undergraduate interns are especially welcome to gain hands-on experience with real systems research and grow into paper or project contributions.
Inha University, Hi-Tech Center
Office 1406 · Student Lab 1404
100 Inha-ro, Michuhol-gu, Incheon 22212, Republic of Korea
인천광역시 미추홀구 인하로 100
인하대학교 하이테크센터
교수실 1406호 · 학생연구실 1404호