What is AI Tagging and Sentiment?
AI Tagging and Sentiment is a powerful Oversai product that uses artificial intelligence to automatically categorize customer interactions and evaluate the sentiment of both the customer and the agent.
This product helps support teams understand the most common reasons customers are reaching out, detect emotional tone in conversations, and identify opportunities to improve the customer experience. It saves time, improves consistency, and adds valuable insights into your support operations.
Step-by-Step: How to Create an AI Tagging and Sentiment Process
1. Navigate to “AI Tagging”
Go to the Logic section in the left-hand menu and click on AI Tagging. Then, click the "New AI Tagging" button in the top-right corner of the screen.
2. Set Up the Basics
Give your AI tagging a name and choose the filters that define which interactions should be analyzed.
🔍 Why Filters Matter
Filters help you narrow down the specific conversations you want to analyze. For example, you might want to tag only closed tickets marked as high priority.
👉 You’ll need to create these filters in the Filters section of the platform before setting up your AI tagging.
Need help? Check out our Filter Creation Guide.
3. Add Relevant AI Tagging Topics
Next, select the topics that matter most to your business. These should represent common reasons your customers reach out—such as:
Product Bug
Login Issues
Delivery Problems
Bad Experience, etc
The AI will automatically assign these tags to conversations based on their content.
Note: You can add up to 25 topics per AI Tagging process, and only one AI Tagging process per integration is allowed.
4. Use the AI Tagging Playground
In the AI Tagging Playground, you can test your setup and see how Oversai’s sentiment algorithm classifies interactions.
Select:
The AI Tagging you want to test
The integration (e.g. Zendesk, Intercom)
The specific interaction ID
Then click Validate.
The system will return the original conversation with AI-generated topics and sentiment scores.
💡 Core Concept: Sentiment Polarity
Sentiment is measured separately for the customer and the agent on a scale from -1 to 1:
-1 = very negative
0 = neutral
1 = very positive
This helps you better understand the emotional tone of conversations and identify potential friction points.
✅ Ready to Launch?
Once you're happy with your test results, contact your Account Manager to activate automated daily analysis.
From that moment, your reports will include accurate AI-generated tagging and sentiment data—fully automated and ready to guide your decisions.
📊 Accessing Reports
To view AI Tagging and Sentiment reports:
Go to the left-hand menu
Click Analytics → Reports
Select AI Tagging
There, you'll find detailed reports with your chosen topics and sentiment scores, helping you track trends and take action faster.
Still need help?
Contact us