Master's thesis Generative AI for the development of embedded device drivers (m/f/d)

Comparative investigation of large language models or code generation models (e.g. GPT, Claude, LLaMA) Development of a “Skill-Layer” (Prompt-Templates, Function Calling, Tool-Use Agents) to control the models specifically for deterministic, safe and verifiable code Development of an end-to-end system that processes design artifacts (switches, data sheets, firmware specifications) converting the artefacts into a model-appropriate representation Driver code (e.g. generated in Python or C++), incl. graphical user interface for integration purposes validated the results by automatic unit tests and hardware-in-the-loop verification

Final result documentation as well as outlook for future developments and possibility of production-ready workflows

Continuous study of electrical engineering, computer science, physics or a comparable degree Ideally basic knowledge in HF measurement technology, signal processing or numerical algorithms Knowledge of how LLMs, RAGs, etc., and software engineering experience in Python or C++ Experience with modern development practices of advantage: Git, CI/CD, Code Reviews, Automated Tests, Clean Code, Windows and Linux Analytical way of thinking, structured working and strong team and communication skills Good knowledge of German in word and writing

Career opportunities Christmas money Coaching offers Business doctor Occupational pension Extensive incorporation Operational childcare Employee discount Fitness facilities Flexible working hours Homeoffice Company events and events Company ticket Further training Holiday allowance The employee benefits shown are company-dependent and may vary depending on the position and department. .