Skip to main content

Quickstart

NaviGator Builder is UF's Pilot AI workflow automation platform, powered by n8n. It lets you visually design workflows that connect apps, services, and AI models without writing large amounts of code. This guide will walk you through logging in, creating your first workflow, and running it.

note

Before you can login to NaviGator Builder, your project must be onboarded and your Gatorlink must be added as a member or admin of the project. See the Project Access page for details on how to request a new project.

Logging In​

  1. Navigate to NaviGator Builder at https://builder.navigator.ai.ufl.edu/.
  2. Sign in with your Gatorlink credentials via UF single sign-on.
  3. After login, select the project you want to work in from the project switcher at the top of the page. If you only belong to a single project, it will be selected automatically.

Creating Your First Workflow​

  1. From the project dashboard, click the New Workflow button in the top right.
  2. Give your workflow a descriptive name by clicking the title at the top of the editor.
  3. You will start with an empty canvas and a Trigger node placeholder. Every workflow in NaviGator Builder starts with a trigger — this is the event that starts the workflow running. Common triggers include:
    • Manual Trigger — run the workflow on-demand from the editor (great for testing)
    • Schedule Trigger — run the workflow on a recurring schedule
    • Webhook Trigger — run the workflow when an HTTP request is received
  4. Click the trigger placeholder and choose Manual Trigger for this first workflow.
  5. Click the + button on the right side of the trigger node to add your next step. You can browse the full catalog of built-in nodes or search for a specific integration.

Adding an AI Step​

NaviGator Builder can call the AI models enabled for your project directly from a workflow step.

  1. After your trigger node, click the + button to add a new node.
  2. Search for HTTP Request (or, if available in your instance, the NaviGator AI node).
  3. Configure the node to make a POST request to the NaviGator AI chat completions endpoint:
    • Method: POST
    • URL: https://api.ai.it.ufl.edu/v1/chat/completions
    • Authentication: choose a credential of type Header Auth with the header name Authorization and value Bearer YOUR_PROJECT_API_KEY
    • Body Content Type: JSON
    • Body:
    {
      "model": "gpt-oss-20b",
      "messages": [
        { "role": "system", "content": "You are a helpful assistant." },
        { "role": "user", "content": "Write a short poem about alligators." }
      ]
    }
  4. Click Execute Node to test the step. You should see the model's response in the output panel.
tip

Credentials in NaviGator Builder are stored encrypted and are scoped to your project. Ask your project admin to create a shared credential so that every workflow in the project can reuse it without exposing API keys.

Running and Saving Your Workflow​

  1. Click Execute Workflow at the bottom of the editor to run the full workflow from the trigger. Each node will show a green checkmark and a preview of its output when it finishes successfully.
  2. Click Save in the top right to save your workflow.
  3. To make the workflow run automatically (for example, on a schedule or webhook), toggle the Active switch in the top right. Only workflows with an automatic trigger need to be activated — manually-triggered workflows can always be run from the editor.

Viewing Execution History​

Every time a workflow runs, NaviGator Builder records the execution. To view history:

  1. Open the workflow in the editor.
  2. Click the Executions tab at the top of the editor.
  3. Select any past execution to replay the data that flowed through each node — this is extremely useful for debugging.

Next Steps​

  • Explore the full node catalog to see the hundreds of built-in integrations available.
  • Add error-handling branches to your workflows so failures can be retried or reported.
  • Work with your project admin to set up shared credentials for the systems your project uses most often.
  • If you need additional AI models enabled for your project, open a ticket with the UFIT Artificial Intelligence & Platform Engineering team.