At FAU, I’m studying Artificial Intelligence and focusing on reliable LLM applications. I built VeritasMed, an LLM-powered question-answering system for medical literature that provides traceable sources.
For VeritasMed, I evaluated retrieval and answer quality with RAGAS and Langfuse, and added source checks, query rewriting, and targeted regeneration to its LangGraph workflow. I also developed the Python data pipelines, FastAPI backend, React frontend, and secured MCP server.
At the Huawei Munich Tech Arena 2026, I independently developed a PyTorch pipeline to locate anatomical landmarks on 3D ear meshes, achieving a 1.23 mm mean Euclidean error against a 5.83 mm baseline. Earlier, as a Software Development Intern at Whale Cloud Computing Co., Ltd., I developed backend features and REST APIs for the Jiangsu Unicom information management project and fixed at least 10 platform issues.
