The Missing Layer Between Enterprise Data and AI Agents

Building one of the largest AI implementations worldwide in the financial sector, Jannis and Sebastian share practical insights into the architectural design and operation of a sovereign Data & AI platform for the Sparkassen Group. They demonstrate how Agentic AI can be built at enterprise scale and operated under strict compliance and regulatory requirements, providing full control over data, models, and infrastructure as a prerequisite for trust, compliance, and scalability. While most discussions focus on models, frameworks, and orchestration, one of the biggest challenges lies in the layer between enterprise data and AI systems: Knowledge and Intelligence as a Service. AI agents are only as effective as the data, context, and operational controls behind them. For years, data debt was mostly a reporting problem. Agentic AI is the first technology that forces organizations to deal with it, bringing data quality issues directly into business processes and customer interactions. In this session, they explore the architectural foundation that enables trustworthy AI at scale: metadata, semantic models, governance, observability, and trusted business abstractions. We show how these components transform fragmented enterprise data into a reliable foundation for AI-powered services, conversational business intelligence, and agentic experiences. Drawing on real-world lessons from building and operating a sovereign Data & AI platform, we demonstrate how organizations can create trustworthy intelligence on top of enterprise data while maintaining full control over data, models, and infrastructure as a prerequisite for trust, compliance, and scalability. This is a practical story about the journey from analytics platforms to intelligent systems and why Agentic AI does not start with agents, but with data, semantics, trust, and knowledge as a service. Attendees will gain practical insights into building scalable AI architectures that balance innovation, compliance, and trust.