Build Your First Workflow

This tutorial walks you through building your first workflow in Fabric AI. In about 15 minutes, you'll create an automated pipeline that processes data with AI and sends results to Slack.

What You'll Build

By the end of this tutorial, you'll have:

• Created a workflow with a webhook trigger
• Added an AI processing node
• Connected a Slack notification node
• Published and tested the workflow end-to-end

Prerequisites

Before starting, make sure you have:

• ✅ A Fabric AI account (sign up here)
• ✅ An AI provider configured (see AI Configuration)
• ✅ A Slack workspace with a bot token (optional — you can skip the Slack step)
• ✅ 15 minutes of time

Step 1: Create a New Workflow

Step 2: Add a Trigger

Every workflow starts with a trigger — the event that kicks off execution.

Step 3: Add an AI Processing Node

Now add an AI node that processes incoming data.

Analyze the following text and provide:
1. A one-sentence summary
2. Key topics mentioned
3. Sentiment (positive, negative, or neutral)

Text: {{trigger.body}}

Step 4: Add a Notification (Optional)

Send the AI's output to Slack.

📊 New Analysis Complete

{{AINode.content}}

Don't have Slack? Skip this step — you can still test the workflow and view results in the execution log.

Step 5: Publish and Test

curl -X POST https://your-webhook-url \
  -H "Content-Type: application/json" \
  -d '{"body": "Our team launched the new dashboard feature this week. Customer feedback has been overwhelmingly positive, with a 40% increase in daily active users. However, we noticed some performance issues on mobile devices that need attention."}'

Step 6: Iterate and Improve

Now that your workflow is running, try these enhancements:

Add Conditional Logic

Insert a Condition node after the AI step:

If {{AINode.sentiment}} equals "negative"
  → Send to #urgent-alerts
Else
  → Send to #general-updates

Add Error Handling

Add an Error Handler node to catch failures and send fallback notifications.

Try AI-Generated Workflows

Instead of building manually, describe what you want:

"Create a workflow that receives customer feedback via webhook,
analyzes sentiment with AI, creates a Linear issue for negative
feedback, and posts a summary to Slack"

The AI generates the entire workflow for you.

Tips for Better Workflows

Keep It Simple

Start with 3-4 nodes and add complexity later. Each node should do one thing well.

Use Template Variables

Reference previous node outputs with {{NodeName.field}}:

{{trigger.body}}        — Webhook payload
{{AINode.content}}      — AI response text
{{GithubNode.issueUrl}} — Created issue URL

Test Incrementally

Run the workflow after adding each new node. This makes it easier to debug issues.

Use Descriptive Names

Rename nodes from defaults ("AI Node 1") to descriptive names ("Analyze Sentiment") for clarity.

What's Next?

Troubleshooting

Workflow doesn't trigger

• Ensure the workflow is Published (draft workflows don't execute)
• Verify you're POSTing to the correct webhook URL
• Check that the request body is valid JSON

AI node returns empty output

• Verify your AI provider is configured in Settings → AI Providers
• Check that the template variables resolve correctly (view the node's input in execution details)

Slack message not sent

• Confirm the bot is invited to the target channel
• Check the Slack node's execution output for error details
• Verify the Bot OAuth Token hasn't expired

Next Steps