I build software where systems meet intelligence — real-time sensor pipelines in robotics, ML-powered prediction models, and mobile products that ship.
B.S. Software Convergence at Kyung Hee University (2021–present). My background spans robotics sensor systems at ROBROS, multimodal ML research, and co-founding a startup that reached the App Store.
Built the image communication layer connecting Intel RealSense D435 and stereo cameras to the ROBROS humanoid platform. Optimized Linux kernel networking and tuned CycloneDDS middleware across heterogeneous LAN/WiFi environments to ensure reliable, low-latency image streams.
Led end-to-end research on the impact of ChatGPT on student learning outcomes — from hypothesis design to model validation. Preprocessed multi-modal datasets and implemented statistical and ML/DL models for the analysis pipeline.
Co-founded a photo studio booking platform for the Korean consumer market. Led a team of 3 engineers and 2 designers from zero to App Store. Established B2B partnerships and drove mobile app monetization through iOS.
Designed a Kubernetes-based LLM inference platform focused on cost-efficiency and operational simplicity. Automated the full deployment lifecycle — CI/CD pipelines, Tailscale mesh networking, and K8s Ingress routing.
Built a system that infers cognitive focus from body signals — HRV, EDA, and skin temperature collected from wearables. Addressed class imbalance and signal resampling to push classification accuracy on an imbalanced real-world dataset.
Open to software engineering roles, research collaborations, and interesting problems.
dlehddnjs245@khu.ac.kr