M.Sc. Computer Science @ TUM | SAP Full Stack Developer | AI & HPC Researcher
I am a Master's student at the Technical University of Munich, combining enterprise-grade software engineering with high-performance computing research. My focus is on building clean architecture, optimizing backend performance, and developing reliable systems.
- Email: yaxuan.chen@tum.de
- LinkedIn: [LinkedIn]
| Domain | Technology Stack |
|---|---|
| SAP Ecosystem | ABAP, RAP (Managed/Unmanaged), CDS Views, SAPUI5 / Fiori, OData (V2/V4), CAP, BTP (Clean Core) |
| Core Engineering | Java (Spring Boot), C++ (HPC/Optimization), Python, SQL, JavaScript, Git, Docker, Linux |
| Research & AI | PyTorch, CUDA, Stable Diffusion, YOLO, pybind11, GitHub Actions (CI/CD) |
Research Project @ TUM Chair of Communication Networks
An open-source, high-performance Reliability Block Diagram (RBD) evaluation tool.
- Contribution: Designed a C++ boolean evaluation engine and integrated it via pybind11 to resolve Python performance bottlenecks.
- Impact: Achieved a 200x performance increase; Core logic accepted at RNDM 2025.
- Tech: Python, C++, Boolean Algebra, HPC Profiling.
SAP Full-Stack Business Application
An end-to-end maintenance management system deployed on SAP BTP, strictly following Clean Core principles.
- Backend: Implemented RAP (Managed) behavior definitions, semantic CDS data modeling, and ETag concurrency control.
- Frontend: Developed a custom SAPUI5 application with XML views and OData V2 binding.
- Tech: ABAP, RAP, CDS, SAPUI5, OData.
Full-Stack Java Application
A distributed flight booking platform featuring complex business logic and user workflows.
- Architecture: Built with Spring Boot (Backend) and JavaFX (MVC Frontend).
- Engineering: Implemented comprehensive Unit Tests (JUnit 5), external API integration, and relational database modeling.
- Tech: Java, Spring Boot, REST API, SQL.
AI Application
A real-time speech-to-image generation pipeline designed for VR environments.
- Pipeline: Integrated OpenAI Whisper for transcription and Stable Diffusion 3.5 for image generation.
- Analysis: Utilized YOLO and PySlowFast for human action and emotion recognition in workplace settings.
- Tech: Python, PyTorch, Multimodal LLMs.
Technical University of Munich (TUM)
- M.Sc. Computer Science (Current)
- B.Sc. Computer Science (Thesis Grade: 1.0)


