SITIC Special Event AI 2026, 22 – 23 June, Schloss Liebegg, CH-5722 Gränichen AG
Where
Business Stakeholders and IT Professionals
Share AI Know-how
Artificial intelligence has long been more than just a topic for the future – it is already being successfully used and further developed today. Many organisations are faced with similar questions:
- Which use cases work in practice? Where does AI bring real added value?
- Why did AI projects fail? Which use cases did not work in practice?
- How do you organise governance over the use of AI?
- And what insights from previous AI projects would you like to share and compare with others in order to secure the next steps and develop new ambitions?
(All images legally acquired)
SITIC invites IT user organisations to present specific AI use cases that have already been successfully implemented, are currently in planning, or even those that were not successful – because, as we all know, we learn the most from our mistakes. Whether initial pilot projects, productive solutions or strategic plans: your practical insight is of great value to other organisations.
Why participating?
Why your participation will benefit your organisation:
- Direct insight into real challenges and solutions
Lessons learned from other organisations implementing AI, including what worked, what didn’t, and why. - Benefit from proven methodologies
Which approaches, frameworks, and tools have been successful in delivering AI applications in other organisations. - Positioning as an active AI leader
Participation demonstrates your organisation’s commitment to innovation and technological advancement. - Constructive feedback and fresh inspiration
Gain new perspectives from peers who are working on similar challenges and opportunities. - Valuable professional connections
Build relationships with business and IT leaders who are making comparable strategic decisions.
The event emphasises real experience, honest lessons learned, and open, collegial exchange on equal footing — not abstract concepts or high‑level plans.
Location
A shuttle service takes public transport users directly from Gränichen to the castle and back in just a few minutes. Interested parties will receive personal assistance.
Language / Sprache
Presentations / Präsentationen
Englisch or German, according to individual preference / Englisch oder Deutsch, je nach individueller Präferenz.
Plenary and Group Discussions / Plenums- und Gruppendiskussionen
Preferably in German / Vorzugsweise in Deutsch.
Participants
Business stakeholders and IT professionals of larger IT user organisations who are responsible for business development, digitalisation, AI application, enterprise and IT architecture, software engineering, compliance and business ethics.
Planning, Organisation, Administration & Moderation
Agenda 22 June, 09:10 – 17:10, plus Apéro Riche
Opening Keynote
Prof. Dr. Stephan Aier, SITIC/University St. Gallen
AI User Presentations, 40 minutes each
(maximum of 12 presentations in parallel tracks)
The date and time of the presentations still need to be agreed upon with the speakers.
Agent Integration Patterns: From API-First to Enterprise Ready Agentic AI
David Bolli, AXA
Agentic AI entfaltet in Unternehmen erst durch nahtlose Integration das volle Potenzial: Die Kollaboration zwischen Agenten und der gezielte Einsatz von Umsystemen als Tools sind dabei essentielle Bausteine.
Diese Präsentation beleuchtet, wie neue Protokolle wie A2A und MCP diese Kollaboration ermöglichen – und welche Herausforderungen bei der Integration von Agenten in bestehende Unternehmenslandschaften zu meistern sind.
AI Governance – Das Fundament für skalierbare KI-Innovation
Christoph Kaufmann, BKW
In dieser Präsentation erhältst du einen Einblick in den strukturierten und ambitionierten Start von KI-Governance bei BKW mit dem klaren Ziel, Innovation zu ermöglichen und gleichzeitig ein starkes Governance-Fundament zu schaffen. Du erfährst, wie klare Leitplanken und ein eng abgestimmtes Innovationsteam Geschwindigkeit, Orientierung und Sicherheit vereinen und wie sich unsere Governance angesichts neuer regulatorischer und technologischer Entwicklungen weiterentwickelt hat.
Von Legacy Systemen zum Agentic Knowledge
Max Wittig, APG|SGA
Während KI zunehmend in Software, Prozesse und Strategien einzieht, entsteht eine neue Architekturfrage: Wie machen Unternehmen ihr Wissen für LLMs und Agenten dauerhaft, nachvollziehbar und sicher nutzbar?
Der Vortrag zeigt anhand aktueller APG|SGA-Vorhaben unseren Weg von bestehenden Altsystemen zu agentenfähigen Architekturen: Vom Aufbau eines AI-Backbone auf Basis von Infrastructure as Code über AI-Gateways bis hin zu ersten PLSQL-Agenten für Legacy-Anwendungen.
In einem zweiten Teil werden zwei explorative Modelle zur Wissensstrukturierung mit KI-Sprachmodellen betrachtet und gemeinsam diskutiert, wie aus Daten, Code und Dokumentation ein belastbares Langzeitgedächtnis für Enterprise AI entstehen kann.
<– More slots available – For your contribution? –>
Four World Cafés, 2 x 70 minutes
1. Governance & Ethics
2. Digital Sovereignty
3. & 4. Democratic Evaluation
Registered participants can submit topic suggestions before the event.
These will be prioritised at the event, and the top two will be put up for discussion.
Agenda 23 June, 09:10 – 17:10, plus Apéro Riche
AI User Presentations, 40 minutes each
(maximum of 12 presentations in parallel tracks)
The date and time of the presentations still need to be agreed upon with the speakers.
Do Cybernetic Developers dream of AI Code?
Francesco De Collibus, AXA
Wir treffen eine neue Art von Mitarbeitern, die kybernetischen Entwickler, eine Mischung von Menschen und Maschinen, von KI und Gehirn. Sie arbeiten für eine neue Organisation: die CybOrg, die Kybernetische Organisation. Diese neue Initiative der AXA Schweiz hat das Ziel, unsere Mitarbeiter systematisch zu befähigen und zu „augmentieren» – zu empowern. Wir wollen nicht nur KI-gestützte Software oder Zeilen von Code, sondern ganzheitliche, KI-gestützte Lösungen schaffen. Wir werden gerne unseren aktuellen Stand präsentieren.
Von KI‑Plattformen zu messbaren Ergebnissen – Erfahrungen aus Transformationsvorhaben
Dr. Marc L. Brogle & Stephan Kreis, DXC Technology
In dieser Präsentation geben wir einen praxisnahen Einblick, wie KI in realen Transformationsvorhaben eingesetzt wird, um komplexe Software‑ und Modernisierungsinitiativen messbar zu beschleunigen. Anhand konkreter Kundenbeispiele – von KI‑gestütztem Knowledge Management mit Copilot Studio, über AI‑Agent‑Plattformen zur Skalierung neuer Systeme bis hin zur Modernisierung großer Legacy‑Landschaften – zeigen wir, wo KI heute bereits echten Mehrwert stiftet. Im Fokus stehen die zentralen Lessons Learned aus diesen Projekten: welche technologischen, organisatorischen und Governance‑Aspekte entscheidend sind, damit KI nachhaltig skaliert und Wirkung entfaltet.
Skalierbare GenAI‑Plattform @ ZKB
Sonja Huclova, Zürcher Kantonalbank
Diese Präsentation zeigt, wie wir eine skalierbare GenAI‑Plattform aufbauen, die heutige und zukünftige Anforderungen an Handlungsfähigkeit und Innovation erfüllt. Wir beleuchten die Ausgangslage, zentrale Architekturentscheidungen sowie technische und organisatorische «Lessons Learned» auf unserem Weg. Darüber hinaus erläutern wir, wie wir die Plattform so gestalten, dass sie neue Use Cases unterstützt und mit sich wandelnden Technologien mitwachsen kann. Im Fokus stehen dabei das Zusammenspiel von Skalierbarkeit, Sicherheit und Flexibilität sowie strategische Weichenstellungen, welche die Plattform langfristig zukunftsfähig machen.
Agentic AI CLI Tool in Software Development
Alessandro Surace, Swiss Re
Agentic AI CLI tools are moving from «better chat in a terminal» to systems that can plan, edit, test, call tools, and sometimes open pull requests with limited human intervention. This session explains how to distinguish chat mode from true agent mode, how to evaluate tools for enterprise use, and which controls architects need across identity, model routing, permissions, MCP integrations, observability, and governance. Using Swiss Re’s internal evaluation material as a base, we compare leading tools and propose a practical architecture for safe adoption.
Claimy: Wie KI die Leistungsfallbearbeitung messbar transformiert
Ionel Onofrescu, PAX
Claimy zeigt, wie generative KI die Leistungsfallbearbeitung messbar schneller, qualitativ besser und wirtschaftlicher macht. Anhand eines real produktiven Systems erläutert Pax Architektur, Governance und Organisation hinter der Lösung. Im Fokus stehen konkrete Effekte, Vorgehen und Learnings.
Von generativer KI bis MLOps – Einblicke in moderne AI-Entwicklung
Babek Dabaghchian, PAX
Der Vortrag gibt einen praxisnahen Überblick über moderne KI-Entwicklung – von klassischen Machine-Learning-Ansätzen bis hin zu generativen Systemen. Dabei werden zentrale Konzepte, Herausforderungen sowie Best Practices beleuchtet, inklusive eines Einblicks in Retrieval-Augmented Generation (RAG). Zudem wird gezeigt, wie unsere Analytics-Plattform Databricks als integrierte Umgebung für Datenverarbeitung, Modelltraining und MLOps eingesetzt werden kann.
<– More slots available – For your contribution? –>
Four World Cafés, 2 x 50 minutes
1. Governance & Ethics
2. Digital Sovereignty
3. & 4.
As evaluated the previous day.
Summaries & Takeaways
Closing Keynote
Prof. Dr. Reinhard Riedl, SITIC/Berner Fachhochschule
Registration
Participants: Email to iva.hollenstein@sitic.org
Presenting Participants: Email to kurt.wehrli@sitic.org
Fees
Participation is free of charge for
– SITIC members as part of their annual membership.
– Non-members from IT user organisations with a presentation (2 seats for both days).
– IT service providers with a presentation jointly with an IT user organisation (3 seats for both days).
For non-members from IT user organisations, one seat costs CHF 750/day.
AI Consensus

EU Artificial Intelligence Act
Article 3: Definitions
(1) ‘AI system’ means a machine-based system that is designed to operate with varying levels of autonomy and that may exhibit adaptiveness after deployment, and that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments;
Schloss Liebegg
AI and Witchcraft?
@ChatGPT, 24 February 2026, 21:55:
What do AI and witchcraft have in common?
Interesting question 🙂
Here are some thought-provoking parallels between AI and witchcraft:
1. They feel like “magic”
Advanced AI systems like ChatGPT can generate text, images, music, or code in seconds.
To someone who doesn’t understand the underlying mathematics and data processing, it can seem almost supernatural—just like witchcraft once did.
2. Specialized or “secret” knowledge
Historically, witchcraft was associated with hidden knowledge accessible only to initiates.
AI is built on complex mathematics, neural networks, and massive datasets that most people don’t see or fully understand.
3. Pattern recognition and prediction
Witchcraft was often used for divination—interpreting signs to predict the future.
AI analyzes patterns in data to make predictions (weather, disease risk, stock trends, recommendations).
4. Fear and fascination
Both inspire a mix of awe and anxiety:
- Excitement about new possibilities
- Fear of misuse or loss of control
- Moral and ethical debates
Societies once feared witches; today, some fear uncontrolled AI.
5. Input → Output ritual
In a playful sense:
- Spell + incantation → effect
- Prompt + model → response
In both cases, the user performs a structured action expecting a specific result.
Philosophically speaking:
Magic is often just technology we don’t yet understand.











