I'm a Game Software graduate from Hongik University, currently in the SSAFY 16th cohort (Data Track). I started out building game clients with Unity (C#) and Unreal Engine 5 (C++), and I like building tools that take repetitive manual work off people's hands. I'm now extending that into data and AI.
- ETRI Research Intern Β· Jul β Aug 2025
- Built a Unity/C# tool to visualize, trim, and label BVH motion-capture data. It loads whole folders at once, maps JSON labels automatically, and plays back frame by frame. Because of it, recording and validation can now run in parallel instead of one after the other.
- Captured 30+ project-specific motion datasets with OptiTrack, Xsens, and StretchSense
- Added object pooling and spline-based randomized paths to a scenario-based VR simulator (UE5/C++)
- Hongik Metaverse Convergence SW Academy Β· Jan β Jun 2025
- Vice team lead of a 6-person team building a VR career-experience simulator in UE5
- Brought in Agile, a Kanban board, and daily scrums, which cut meetings from 30 min to 5β10 min
- Exhibited at 2025 Global Game Challenge (GGC); co-authored a paper at ICONI 2025
- AI / LLM β Studying LLM fundamentals (SFT, RLHF/DPO, decoding strategies) and VLM fine-tuning
- Led a 4-person team in the SSAFY AI Challenge (image multiple-choice VQA): fine-tuned a Qwen-based VLM with LoRA, raising accuracy from 90% (zero-shot) to 96%
- Data Engineering β Learning event streaming and stream processing with Apache Kafka and Apache Flink (PyFlink), plus Docker
- Web β Built LocalHub, a Seoul travel-spot explorer on Korea's public TourAPI with a GPT chatbot (Vue.js + FastAPI). I owned the frontend, CRUD API design, and most of the deployment.
- Algorithms β Solving problems weekly in C++ and reviewing each other's code via PRs in a 5-person study group

