1 minute read

A few weeks ago, Markus Eisele & Bruno Meseguer gave a great talk with the catchy title “Enterprise Integration Is Dead! Long Live AI-Driven Integration with Apache Camel” at DevoxxUK.

In this talk, they explain that integration isn’t dead at all. It is hiding nowadays behind modern abstractions. They also explain why we still need, and will still need enterprise integration, the challenges we face now and in the future.

They also explain how enterprise integration evolves with AI:

  • AI-driven applications now interpret intent, not just accept inputs
  • Systems can reason about business logic and select integration paths dynamically
  • Integration logic is no longer hardcoded, it’s shaped in real-time

They introduce Apache Camel as an integration solution, with a nice AI-driven integration demo featuring Apache Camel, LangChain4j, Kaoto, Docling.

The demo showcases powerful data ingestion powered by both Apache Camel and Docling. The demo also showcases autonomous agents and agentic integration AI, leveraging Apache Camel Routes as tools, powered by LangChain4j. Additionally, the demo showcases how to create AI-driven workflows with planners and human-in-the-loop capabilities.

They conclude with their vision of the future and how this can shape AI business cases, providing recommendations on how to move from legacy to AI-driven integrations.

They also introduce Wanaku, an open-source project we kicked off recently, powered by Apache Camel and Quarkus as an MCP Router for everything Camel.

I absolutely love how Markus introduced Apache Camel so briefly, and I’m a huge fan of Apache Camel demos from Bruno.

Check out their video now.

You can explore the complete demo implementation in the GitHub repository.

Bruno also provided a tutorial on using tools with Apache Camel on OpenShift.

I’m so happy to see folks present some of the work I’ve been working on, and collaborating on in the Apache Camel community.

There are many exciting developments coming to Apache Camel around LangChain4j. We are working on improving the AiAgent story by introducing AIServices, and will definitely work on enabling agents to communicate with each other, so stay tuned.

Have other ideas? Feel free to reach out.

Comments