When should an AI explain, ask a question or offer a different perspective? We develop adaptive dialogue systems and study how these choices affect understanding.

The goal is AI that helps people think through complex questions. That makes understanding and critical reflection central to how we build and evaluate language models.

• Conduct research on conversational AI and coordinate selected research activities within CREST-AI together with Prof. Christina Niklaus.
• Develop grounded, adaptive dialogue systems using LLM post-training and reinforcement learning.
• Build and evaluate a conversational AI prototype through simulations and user studies.
• Contribute your own research ideas to CREST-AI, publish at leading NLP and AI venues, and present at international conferences.
• Supervise master’s and bachelor’s theses and student projects.
• Support research coordination and academic service.

• PhD in computer science, computational linguistics or a related field, completed by the start date.
• Research experience in NLP, particularly large language models, dialogue systems or natural language generation.
• Strong Python and machine learning skills; experience with LLM post-training, reinforcement learning or user studies is a plus.
• A publication record demonstrating original contributions to NLP, AI or a related field, appropriate to your career stage.
• Ability to work independently, take initiative and collaborate effectively within a research team.
• Enthusiasm for interdisciplinary collaboration, particularly with researchers in philosophy and law; no prior expertise in these fields is required.
• Excellent written and spoken English.

The SNSF-funded CREST-AI project

Help shape a new generation of language AI that adapts to people's understanding. CREST-AI brings together NLP, philosophy and law to explore how AI can help people understand, question and reason about complex information, using legal language as a challenging testbed. You will join the research team led by Prof. Christina Niklaus at the University of St.Gallen and develop adaptive dialogue systems that tailor explanations to users’ understanding.

What we offer

You will work closely with Prof. Christina Niklaus and her team, with room to contribute your own ideas within the project’s research agenda and strengthen your research profile. A competitive salary, funded conference participation and access to project-funded compute time support your work. You will collaborate with researchers at St.Gallen and the University of Zurich, alongside a fellow postdoc developing interpretable models of users’ knowledge and understanding.

Application

Please apply through the HSG online application portal, submitting a single PDF containing: (1) a cover letter, (2) your CV with publication list, (3) one representative publication or preprint, (4) contact details of two referees, and (5) your PhD certificate or confirmation of expected completion.

Review of applications will begin on 23 November 2026 and continue until the position is filled.