From Prototype to Production: Building Real-World AI and MCP Workflows

Generative and agentic AI are moving from demos into production, but building reliable, scalable, and effective workflows is still a challenge. This session offers a grounded, real-world approach to integrating real data across AI and agentic workflows.

Learn how Reno Sun from the Town of Qualicum Beach and Oliver Morris from Avineon-Tensing have been applying AI across a range of FME workflows, starting with generative AI for document extraction and progressing into agentic AI through the FME Flow MCP server. Their work spans practical integration challenges, experimentation with MCP services for specific use cases, and testing with popular AI tools.

Join us to learn how to:

• Apply generative AI to document extraction within an FME workflow
• Use the FME Flow MCP server to bring agentic AI into enterprise and spatial data work
• Build custom MCP services for targeted use cases, such as Survey123 record creation
• Identify where AI delivers real value, where it falls short, and lessons learned from using real municipal data

For organizations exploring AI integration or looking for practical MCP inspiration, this session offers a real applied view of where FME sits in modern AI and agentic workflows.