Skip to main content
Fine-tuning helps you improve your chatbot’s responses over time. When your chatbot gives an incorrect or incomplete answer, you can provide corrections that it will learn from.

Why Fine-Tune?

Even with good training data, your chatbot may:
  • Miss nuances in certain questions
  • Provide outdated information
  • Use the wrong tone or format
  • Lack context for specific scenarios
Fine-tuning addresses these issues by letting you teach the chatbot exactly how to respond.

Using Updated Answers

The Updated Answers feature lets you correct specific responses:
1

Go to Training & Content

Click Training & Content in the sidebar.
2

Select Updated Answers

Click on Updated Answers.
3

Add a correction

Click Add Updated Answer.
4

Enter the question

Type the question or phrase that triggered the wrong response.
5

Provide the correct answer

Write the response you want your chatbot to give.
6

Save

Click Save to apply the correction.
Updated answers interface

Best Practices

Review Conversation Logs

Regularly check your chatbot’s conversations in the Inbox to identify:
  • Frequently asked questions with poor answers
  • Topics where the chatbot seems confused
  • Missed opportunities to provide helpful information

Be Specific

When adding updated answers:
  • Match the exact phrasing visitors use
  • Include common variations of the question
  • Provide complete, helpful responses

Test Your Changes

After fine-tuning:
  • Use Test Agent to verify the improvement
  • Monitor real conversations to confirm it’s working
  • Iterate if needed

Adjusting AI Instructions

For broader changes, update your AI instructions in Settings:
  • Tone: Make responses more formal or casual
  • Length: Set preferred response length
  • Behavior: Define how to handle certain topics
  • Restrictions: Limit what the chatbot will discuss

Ready to automate your customer service with AI?

Join over 1000+ businesses, websites and startups automating their customer service and other tasks with a custom trained AI agent.
Last modified on January 22, 2026