How can technical documentation be transformed into an intelligent, agent-based AI system?

In episode 45 of the “Knowledge Graph Insights” podcast, Max Gärber, Technical Consultant at PANTOPIX, talks to Larry Swanson about the successful development of a system architecture that combines agentic AI with a structured knowledge graph.

Using the ZEISS Service Copilot project as an example, Max Gärber explains in practical terms how modern large language models (LLMs), semantic technologies, and professional knowledge management interact to create real added value for service, engineering, and sales.

This cannot be achieved with artificial intelligence alone. It requires a stable knowledge graph, clean metadata structures, a well-thought-out ontology and taxonomy architecture, and established knowledge management processes.

This episode therefore also revolves around the question of why many AI projects fail. The conversation makes it clear that successful AI strategies do not start with large language models, but that clean information architecture, semantic modeling, structured content, and sustainable knowledge management play an essential role.

Listen in if you want to use AI in technical documentation, service, or industrial environments. The podcast with Max Gärber provides valuable insights from practice – from strategic conception to concrete system architecture.