Auto Labeling

Video Tutorial

Overview
Auto Labeling is an integral part of the integrations in Seasalt.ai’s suite of tools. This feature intelligently and automatically applies labels to conversations, streamlining processes for use cases such as customer service and healthcare.

We welcome you to check out our feature of Label Automation for more advanced workflows as labe-automation builds upon the auto Labeling feature.

What Is Auto Labeling?
Auto Labeling is a system that automates the application of labels to conversations or other data points based on predefined rules and AI-driven analysis. Instead of manually tagging, users can set conditions under which labels are applied or removed dynamically.

Key Features

Overview of the Auto Labeling UI

  1. Automated Label Application

    • Adds labels like Escalated or Low Priority based on conversation content.
    • Handles multiple labels simultaneously.

  2. Dynamic Rule-Based System

    • Users can define triggers for labels, such as specific keywords or phrases.
    • Includes functionality to both add and remove labels automatically.

  3. Custom Use Cases

    • Examples include healthcare hotlines, where Escalated might apply to severe pain, or Low Priority to minor aches.

  4. Visual Feedback

    • Labels appear in conversation histories, providing instant insight into their classification.

How to Set Up Auto Labeling

  1. Navigate to Integrations
    • Open the Integrations section and select Auto Labeling.
Navigation to Auto Labeling

  1. Create Rules
    • Define triggers for adding or removing labels.
    • Example: For healthcare hotlines:
      • Extreme pain triggers the Escalated label.
      • Minor aches triggers the Low Priority label.

Define Labeling Rules
  1. Test Your Rules
    • Simulate conversations to see if labels apply correctly.
    • Example: Input “I feel so much pain, I need a doctor’s attention” to trigger Escalated.
Test Labeling Rules in Conversation

  1. Adjust and Save
    • Modify rules to refine accuracy and save the configuration.
Define Labeling Rules

Advanced Use Cases

  1. Medical Triage System
    • Labels like Green, Yellow, and Red can classify conditions based on severity.
      • Red: Severe issues like difficulty breathing or severe bleeding.
      • Yellow: Moderate pain or fever.
      • Green: Minor aches or colds.
Define Labeling Rules

  1. Label Automation
    • Link actions to specific labels:
      • Trigger an email or SMS when Escalated is applied.
      • Automate follow-ups for Low Priority cases.
    • Check out our tutorial on Label Automation for more details.

Benefits of Auto Labeling

  • Efficiency: Reduces manual labeling efforts.
  • Scalability: Handles high volumes of data.
  • Accuracy: Ensures consistent application of rules.
  • Customization: Adapts to various industries and use cases.

Conclusion
Auto Labeling transforms how businesses manage conversations and data. From dynamic triaging in healthcare to streamlining customer support, this tool simplifies and accelerates processes while ensuring accuracy. Future tutorials will delve into removing labels and leveraging label automation for more advanced workflows. Stay tuned!

← Live Agent
Long-Term Memory →