Adoption is the hard part
What it takes to get non-developers using agents on real company data: champions, demos, support, trust, and the honest adoption numbers behind the launch.
AI/MLMCPEnterprise AI
Connect
If something here sparked a question (about MCP architecture, enterprise AI, or anything at the intersection of engineering and AI), I'm happy to talk.
Speaking
Available for talks on production MCP and enterprise AI
I'm open to conference talks, meetups, podcasts, and technical sessions about the practical side of agentic systems: adoption, metadata, and MCP Apps.
Upcoming events where I'll be speaking
No public speaking dates are listed yet.
Speaking topics
Domain knowledge belongs in the MCP server
How rich tool descriptions, schemas, and expert-validated metadata turn thin API wrappers into MCP tools an agent can actually reason with.
MetadataTool descriptionsDomain knowledge
The agent is the new browser
How MCP Apps change frontend work: compact agent-native views, metadata-guided rendering, and controlled app-only actions for safer interaction.
MCP AppsAgent UIJavaScript
Send a message
The form below drops straight into my inbox. LinkedIn and email work just as well.
Based in the Netherlands.