Nicolas Sebastian Schuler
PhD student at Karlsruhe Institute of Technology (KIT)
I study how software systems can adapt to changing conditions while remaining dependable. My doctoral research treats the points where learned components hand results to planners, monitors, and policies as architectural seams: specified, checked at runtime, and able to evolve.
Research questions
Software needs to respond when its environment changes or something goes wrong. I explore how to make those responses useful, understandable, and checkable.
My central question: when a learned model passes a prediction, label, or text to a rule-based component, which assumptions can that component rely on, and how can that boundary be specified, monitored, and revalidated as both sides change?
Two related studies look at reasoning in software: whether language models preserve how strongly a stakeholder meant a requirement, and how an image explanation can become logical facts that formal reasoning can check. Alongside this, I have worked on which crash evidence helps a language model repair a program and on evaluating explanations of federated models.
Each question has a paper behind it; see my selected papers. For thesis and project enquiries, see teaching and supervision.
For software and supporting material, explore my projects.
News
- Best Paper Award at the ECSA 2026 Doctoral Symposium (Bolzano) for “Architecting Self-Adaptive Systems with Learned and Symbolic Components”
- TraceGate tool paper accepted at ASE 2026 Tools and Datasets
- TraceGate accepted at the ISSRE 2026 Research Track
- Preprint: Metric-Guided Attribution Fusion in Explainable Federated Learning (submitted to Information Fusion)
- “Bridging Explanations and Logics” accepted at AISoLA 2025
On X
@NSchuler01Read my posts on X.
Away from research
Outside of work, I’m learning 🇯🇵 Japanese and enjoy riding my 🏍️ motorcycle. Coffee and pizza are usually welcome, too.
If you’d like to talk about research, software, or a shared interest, get in touch.
Best regards,
Nicolas