Quickstart

Overview

In this quickstart guide, you’ll build a Camel route that automatically summarizes text files using a local AI model. The route will monitor a folder for new text files, send their content to Ollama (running the Granite AI model), and output concise summaries.

Prerequisites

Before starting, ensure you have:

• VS Code installed on your system - Follow the installation guide
• Kaoto VS Code extension - See the installation guide for setup instructions
• Ollama installed and running - Follow the Ollama installation guide to set it up
• Granite 4 model pulled in Ollama - Run ollama pull granite4:tiny-h in your terminal (https://ollama.com/library/granite4)

Create and Run Your First Route

Navigate to Kaoto Perspective

1. Open VS Code
2. Click on the Kaoto icon in the Activity Bar (left sidebar)
3. The Kaoto perspective will open, showing the route designer

Create a New Camel Route

1. In the Kaoto perspective, click “Camel File…”

   

2. Choose “Camel Route” as the route type

   

3. Choose “YAML DSL” as the route type

   

4. Give a name to the new route

   

5. A basic timer-based route will be created automatically

Image Preview View YAML

View Image YAML Source

- route:
    from:
      uri: timer:yaml
      parameters:
        period: "1000"
      steps:
        - setBody:
            simple: Hello Camel from ${routeId}
        - log: ${body}
    id: route-1905

Run the Simple Route

1. Click the “Run: Workspace” button in the Integrations panel

   

2. The route will start executing, check the terminal output - you should see “Hello Camel from route1” printed every second

   

3. To stop the route, you can click on the “Terminate” button

   

Add AI Summarization

For this route, we’re gonna use Apache Camel 4.18 (or newer), make sure to select that version in the Runtime selector

For starters, instead of automatically producing empty messages with the timer component, we’re gonna use a file component for reading text files from a directory where we’re gonna place the text to summarize

Add the File Component

1. Hover over the timer node and click the replace button

   

2. In the component catalog, search for “file” and select it

   

3. You’ll end up having something like this, notice how hovering over the exclamation mark tell us that we’re missing the directoryName configuration

   

4. Click on the file component, and fill the directoryName field with ./input

   

5. Click on the All toggle, and search for the noop property and enable it

   

6. Search for the idempotent property and enable it

   

7. Hover over the setBody node and click the delete button

   

8. Switch to the File explorer

   

9. Create a new folder called input with a story.txt file inside where we’re gonna write the text to summarize.

   

10. Navigate back to the Kaoto perspective and run the route, you should get a similar output like this. Notice how we can see the text as the route output.

Image Preview View YAML

View Image YAML Source

- route:
    from:
      uri: file
      parameters:
        directoryName: ./input
        idempotent: true
        noop: true
      steps:
        - log: ${body}
      id: from-6531
    id: route-1905

11. We can stop the route now. In the next step, we’re gonna add the open ai component to perform the summary

    

Add the Open AI Component

For the summary process, we’re gonna use the Open AI component to leverage our local Ollama instance with the Granite model

1. Hover in the edge between the file and log components, and click on the Add step button

   

2. In the component catalog, search for openai

   

3. Once the openai component is added, click on it and pick the All view to configure the follow options:

   

Tip

After editing the option, you can click on the Modified toggle to see what has changed

4. Time of truth, let’s run the route and see the summary. If everything went well, you should get a similar output like this:

2026-03-26 13:09:37.293  INFO 592754 --- [ file://./input] demo.camel.yaml:19                       : Once upon a time in Wonderland, Alice dreamt that her childhood fascination turned into a reality where she met her favorite characters from "Alice in Wonderland," but this time it's all around her, filling her ears and eyes! Oh, and let's not forget about the sneezing pig-baby and execution-ordered rabbit. Truly an unforgettable experience!

(Note: This is a lighthearted reimagining of the original text.)

Image Preview View YAML

View Image YAML Source

- route:
    from:
      uri: file
      parameters:
        directoryName: ./input
        idempotent: true
        noop: true
      steps:
        - to:
            id: to-2057
            uri: openai
            parameters:
              operation: chat-completion
              baseUrl: http://localhost:11434/v1
              systemMessage: Summarize this text into a couple of short sentences and give it
                a funny twist
              apiKey: OLLAMA-LOCAL-KEY
              model: granite4:tiny-h
        - log: ${body}
      id: from-6531
    id: route-1905