Johanes Andrian Kartono

Development of an AI agent for SCPI code generation using LangChain, enabling automated operation and testing of measurement devices. Implementation of MCP servers for SCPI command retrieval via RAG and direct device execution. Design of a performance evaluation framework using MLFlow and Pytest, with iterative optimization through testing and analysis.

Development of internal Python tools for test automation (changelog generator, test suite verdict calculator) and creation of Docker containers for reproducible CI/CD pipelines. Management of development infrastructure using Jenkins, GitLab CI, Bash, and PowerShell on Windows and Linux. Advisory support for internal teams on containerization and tool development, alongside maintenance of technical documentation.

Implementation of fuzzing techniques for hardware design verification using Verilator. Conducted literature research on state-of-the-art hardware fuzzing methods to identify potential research gaps. Enhanced the functionality of the Cascade fuzzer through integration of Large Language Models (LLMs). Containerization of all required tools and dependencies using Docker for a reproducible development environment.

Design and configuration of a Hardware-in-the-Loop (HiL) test bench for validating communication protocols of automotive control units. Analysis of vehicle network traces across multiple protocols (CAN, LIN, FlexRay, Automotive Ethernet). Development of automated solutions for test bench setup and maintenance using Python, Bash, and PowerShell. Configuration of VLANs for device separation and segmentation within automotive network environments.

Program Link: https://www.cit.tum.de/cit/studium/studiengaenge/master-elektrotechnik-informationstechnik/

Program Link: https://www.cit.tum.de/cit/studium/studiengaenge/bachelor-elektrotechnik-informationstechnik/