PhD in Trustworthy LLM-Enabled Systems: Requirements and Architecture

This PhD position at Eindhoven University of Technology (TU/e) focuses on designing trustworthy systems that integrate large language models (LLMs) into collaborative environments involving people, organizations, and AI services. The research bridges requirements engineering, software architecture, and trustworthy AI, addressing critical questions about trust dynamics—such as who trusts whom, what they trust, and under what conditions. The candidate will develop methods to translate abstract trustworthiness principles (e.g., transparency, accountability) into concrete architectural decisions while navigating trade-offs like openness versus privacy or automation versus human control. Embedded in the LLM4LM project—a collaboration between TU/e, TNO, and industry—the role offers hands-on experience with real-world AI adoption and decision-making across logistics and regulatory compliance domains. The empirical and design-oriented approach combines qualitative methods (interviews, workshops) with model development, allowing the PhD student to shape their research trajectory within a multidisciplinary team.

• Conduct empirical research to analyze how stakeholders perceive trust in LLM-enabled systems and translate these perceptions into actionable requirements.
• Develop and evaluate methods or models for identifying trustworthiness requirements and guiding architectural decisions in complex systems.
• Collaborate with academic and industry partners to study real-world AI adoption and architectural decision-making processes.
• Design and implement case studies or prototypes to test and refine proposed approaches for trustworthy AI systems.
• Present research findings at conferences, workshops, and internal project meetings.
• Contribute to teaching and mentoring activities, including coaching students and developing teaching skills.