Talks
Talks and keynotes on agentic engineering, software factories, and how organizations learn from working with AI.
Available in German and English for conferences, meetups, and company events.
Current talks
From harness engineering to a software factory
How does a product team organize its work when agents can build substantial parts of the software? This talk approaches the software factory as an operating model the whole team owns and improves. We look at the feedback agents need, how to evaluate results against product goals, and how lessons from daily work improve the factory itself.
For engineers, architects, and product teams.
Read the talk · Read the article
Everyone has AI. The company learns nothing.
People are finding useful ways to work with AI. How much of that experience reaches the next person or team? This talk explores how organizations can retain what they learn, make useful practices available to others, and understand what changed because they invested in AI.
For leaders, product teams, and people responsible for AI adoption.
Read the talk · Read the article
The Elastic Loop: introducing agentic engineering strategically
How much work can you delegate to an agent before you need to look again? The Elastic Loop helps teams choose how closely to collaborate with agents, what feedback to provide, and how to judge the result. It’s about making room for more ambitious ideas and better software.
For everyone working on software products.
Read the talk · Explore the agent-ready book
Also available
Integrating AI agents with MCP
Connecting agents to existing software and data through the Model Context Protocol, with a focus on distributed enterprise systems.
Archive
- Harness engineering: the pressure reactor of agentic engineering
- RAG: The architecture of reliable AI (also available for reading as a Talk+)
- High agency as a survival strategy? How AI is reshaping engineering organizations
- Stop „trying out” AI – Start shipping systems
- Bicycle of the mind – How AI is changing my learning
- Stagnation in progress: The paradox of how we deal with AI (also available for reading as a Talk+)