AI Agent: Analyze Typeform Feedback Sentiment & Notify on Mattermost

Unlock Instant Insights from Your Feedback with this AI Agent

This n8n workflow acts as an AI agent that connects directly to your Typeform. When a new response is submitted, it extracts the relevant feedback, uses Google Cloud Natural Language to determine its sentiment (positive, negative, or neutral), and then intelligently routes this information. If the sentiment meets your defined criteria, it posts a detailed message to a designated Mattermost channel, allowing your team to stay informed and react promptly.

Key Features & Benefits

• Real-time Feedback Processing: Connects directly to Typeform to process submissions as they arrive.
• AI-Powered Sentiment Analysis: Leverages Google Cloud Natural Language for accurate understanding of feedback tone.
• Instant Team Notifications: Automatically alerts your team on Mattermost about significant feedback.
• Customizable Alert Logic: Define specific sentiment scores or conditions to trigger notifications.
• Improved Responsiveness: Quickly identify and address customer concerns or celebrate positive input.
• Time Saver: Automates the manual process of reviewing feedback and sharing it internally.

• An n8n instance (Cloud or self-hosted).
• Typeform account and API credentials.
• Google Cloud Platform account with the Natural Language API enabled and credentials (OAuth2 setup in n8n for the 'Google Cloud Natural Language' node).
• Mattermost instance and API credentials (or bot token) with write access to the target channel.

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Configure the 'Typeform Trigger' node:
   • Select or create your Typeform API credentials.
   • Enter the Form ID of the Typeform you want to monitor. You can find this in your Typeform URL.
4. Configure the 'Google Cloud Natural Language' node:
   • Select or create your Google Cloud Natural Language OAuth2 API credentials.
   • Verify the Content field expression: ={{$node["Typeform Trigger"].json["What did you think about the event?"]}}. This must correctly point to the text field in your Typeform response that contains the feedback. Adjust the JSON path (e.g., {{$json.form_response.answers.find(a => a.field.ref === "your_field_ref_here").text}} or similar) based on your Typeform's structure and the specific question you want to analyze.
5. Configure the 'IF' node:
   • The existing condition {{$node["Google Cloud Natural Language"].json["documentSentiment"]["score"]}} needs to be completed. Define the logic for when to send a Mattermost message. For example, to act on positive feedback (score > 0.2):
     • Set 'Value 1' to {{$node["Google Cloud Natural Language"].json["documentSentiment"]["score"]}}.
     • Set 'Operation' to 'Larger Than'.
     • Set 'Value 2' to 0.2.
   • Adjust as needed for negative sentiment (e.g., 'Smaller Than' -0.2) or other thresholds.
6. Configure the 'Mattermost' node:
   • Select or create your Mattermost API credentials.
   • Update the Channel ID to your desired Mattermost channel where notifications should be posted. You can get this from the channel's URL or settings.
   • Customize the Message content using available data from previous nodes (e.g., feedback text, sentiment score).
7. (Optional) Customize the 'NoOp' node: This node is currently a placeholder for the 'false' branch of the IF statement (i.e., when the feedback doesn't meet your notification criteria). You can replace it with other actions or remove it if no action is needed.
8. Activate the workflow.