Skip to content
Workflows

Workflow Action - AI Intent Detection

The AI Intent Detection action analyzes text to automatically determine sentiment, categorizing it as POSITIVE, NEGATIVE, or NONE. This premium action…

3 min read697 words4 explained images
View as markdownOpen in ClaudeOpen in ChatGPT
Each image has a one line explanation. Switch to full detail for the step it belongs to, the fields and buttons involved, examples and the whole procedure.
On this page
  1. Step 1: Add the AI Intent Detection Action
  2. Step 2: Configure the Action
  3. Step 3: Set Up Branch Actions
  4. 1.Customer Feedback Management
  5. 2.Post-Purchase Follow-up

The AI Intent Detection action analyzes text to automatically determine sentiment, categorizing it as POSITIVE, NEGATIVE, or NONE. This premium action replaces the legacy intent detection in IF/Else conditions, offering improved accuracy and easier implementation for automating responses based on customer sentiment.

TABLE OF CONTENTS


What is AI Intent Detection?

The Intent Detection action is an AI-powered workflow tool that automatically analyzes text input to determine the underlying sentiment or intent. It processes any text—whether from customer messages, survey responses, or form submissions—and categorizes it into three distinct branches:

  • POSITIVE: Text expressing positive sentiment, satisfaction, or agreement
  • NEGATIVE: Text conveying negative sentiment, dissatisfaction, or disagreement
  • NONE: Text with neutral sentiment or when the system cannot confidently categorize it

This action creates three workflow branches automatically, allowing you to build different automation paths based on the detected sentiment.

Key Benefits

  • Automated Sentiment Analysis: Instantly categorize customer responses without manual review
  • Improved Accuracy: More reliable than legacy IF/Else intent detection methods
  • Simplified Workflow Design: Three clear branches eliminate complex conditional logic
  • Real-Time Response: Trigger appropriate actions immediately based on customer sentiment
  • Enhanced Customer Experience: Deliver personalized responses tailored to customer emotions
  • Scalable Automation: Process unlimited messages without human intervention

How to Set Up AI Intent Detection

Step 1: Add the AI Intent Detection Action

Navigate to your workflow builder and search for "AI Intent Detection" in the actions menu. Add it to your workflow at the point where you want to analyze text sentiment.

Step 1: Add the AI Intent Detection Action (image 1 of 4) What this shows Navigate to your workflow builder and search for "AI Intent Detection" in the actions menu. What this shows Illustrates the "Step 1: Add the AI Intent Detection Action" section of "Workflow Action - AI Intent Detection". This screenshot appears in the "Step 1: Add the AI Intent Detection Action" section of "Workflow Action - AI Intent Detection". The text alongside this image reads: Navigate to your workflow builder and search for "AI Intent Detection" in the actions menu. Add it to your workflow at the point where you want to analyze text sentiment. Immediately after, the guide continues: Action Name: Give your action a descriptive name (e.g., "Analyze Feedback Sentiment") Input Text: Specify the text to analyze. You have two options. Image 1 of 4 Where to goNavigate to your workflow builder and search for "AI Intent Detection" in the actions menu. Add it to your workflow at the point where you want to analyze text sentiment. Next stepAction Name: Give your action a descriptive name (e.g., "Analyze Feedback Sentiment") Input Text: Specify the text to analyze. You have two options:

Step 2: Configure the Action

  1. Action Name: Give your action a descriptive name (e.g., "Analyze Feedback Sentiment")
  2. Input Text: Specify the text to analyze. You have two options:
    • Static Value: Enter fixed text directly
    • Dynamic Variable: Use the custom value picker to select variables:
      • {{message.body}}for incoming messages
      • Any other variable that stores text to be classified
Step 2: Configure the Action (image 2 of 4) What this shows Action Name: Give your action a descriptive name (e.g., "Analyze Feedback Sentiment") Input Text: Specify the text to analyze. What this shows Illustrates the "Step 2: Configure the Action" section of "Workflow Action - AI Intent Detection". This screenshot appears in the "Step 2: Configure the Action" section of "Workflow Action - AI Intent Detection". The text alongside this image reads: Action Name: Give your action a descriptive name (e.g., "Analyze Feedback Sentiment") Input Text: Specify the text to analyze. You have two options. This part of the guide covers 2 fields, listed below. Immediately after, the guide continues: After saving, the action automatically creates three branches. Image 2 of 4 What to fill inAction Name: Give your action a descriptive name (e.g., "Analyze Feedback Sentiment") Input Text: Specify the text to analyze. You have two options:
Fields in this part of the guide Static Value: Enter fixed text directlyAny other variable that stores text to be classified
Buttons and menus referenced Action NameInput TextStatic ValueDynamic Variable
Next stepAfter saving, the action automatically creates three branches:
All 2 steps in this procedure
  1. Action Name: Give your action a descriptive name (e.g., "Analyze Feedback Sentiment")
  2. Input Text: Specify the text to analyze. You have two options:

Step 3: Set Up Branch Actions

After saving, the action automatically creates three branches:

Step 3: Set Up Branch Actions (image 3 of 4) What this shows After saving, the action automatically creates three branches: What this shows Illustrates the "Step 3: Set Up Branch Actions" section of "Workflow Action - AI Intent Detection". This screenshot appears in the "Step 3: Set Up Branch Actions" section of "Workflow Action - AI Intent Detection". The text alongside this image reads: After saving, the action automatically creates three branches. This part of the guide covers 8 fields, listed below. Immediately after, the guide continues: 1.POSITIVE Branch: Configure actions for positive responses. Image 3 of 4 What this coversAfter saving, the action automatically creates three branches:
Fields in this part of the guide Send thank you messagesRequest app store reviewsOffer loyalty rewardsTrigger support escalationSend discount codes or couponsSchedule follow-up callsSend clarification requestsRoute to manual review
Next step1.POSITIVE Branch: Configure actions for positive responses
All 4 steps in this procedure
  1. Send thank you messages
  2. Request app store reviews
  3. Offer loyalty rewards
  4. Add "satisfied customer" tags

1.POSITIVE Branch: Configure actions for positive responses

  • Send thank you messages
  • Request app store reviews
  • Offer loyalty rewards
  • Add "satisfied customer" tags

2.NEGATIVE Branch: Configure actions for negative responses

  • Trigger support escalation
  • Send discount codes or coupons
  • Schedule follow-up calls
  • Add "needs attention" tags

3.NONE Branch: Configure actions for neutral/unclear responses

  • Send clarification requests
  • Route to manual review
  • Apply default follow-up sequences

Example Use Cases

1.Customer Feedback Management

Scenario: After collecting feedback through surveys

  • POSITIVE Response: "Your service is amazing!"
    • Action: Send app store review request link
  • NEGATIVE Response: "Very disappointed with the experience"
    • Action: Send 20% discount coupon for next purchase
  • NONE Response: "Okay" or "N/A"
    • Action: Send follow-up questions for more details

2.Post-Purchase Follow-up

Scenario: After delivery confirmation

  • POSITIVE: Request product review and testimonial
  • NEGATIVE: Trigger return/refund workflow
  • NONE: Send product usage tips

Pricing

  • Cost: 1 cent ($0.01) per execution
  • Part of Premium Actions suite

Frequently Asked Questions

Q: How is this different from the legacy IF/Else intent detection?

A: The Intent Detection action uses advanced AI for more accurate sentiment analysis, provides automatic branching, and eliminates the need for complex conditional logic setup.

Q: Can I use multiple Intent Detection actions in one workflow?

A: Yes, you can chain multiple Intent Detection actions to analyze different text inputs or create multi-level sentiment analysis workflows.

Q: What happens if the text contains mixed sentiments?

A: The system analyzes the overall dominant sentiment. If no clear sentiment is detected, it routes to the NONE branch.

Q: Can I test the intent detection before going live?

A: Yes, use the workflow test mode with sample text inputs to verify the sentiment detection accuracy before activating your workflow.

Q: Can I see why text was categorized a certain way?

A: Currently, the action provides the categorization result only. Detailed reasoning or confidence scores may be added in future updates.

Was this guide helpful?

Related guides