Bot Behavior Settings

Bot Behavior settings control AI behavior at the platform level. These settings act as defaults that can be overridden by individual persona configurations.

Accessing Bot Behavior

1. Go to Settings
2. Click Bot Behavior (or find in AI Settings)

Default Response Settings

Response Language

Set the default language for AI responses:

• Auto-detect - Match customer's language
• English - Always respond in English
• Other - Select specific language

Response Length Default

Set default verbosity:

Personas can override this setting.

Markdown Formatting

Allow markdown in responses:

• Enabled - AI can use headers, lists, bold
• Disabled - Plain text only

Markdown improves readability for complex answers.

Citation Settings

Show Sources

Whether to display source citations:

• Always - Show on every response
• When Available - Show when sources are used
• Never - Don't show citations

Citations build trust by showing where answers come from.

Citation Format

How sources are displayed:

• Inline - Referenced within text
• Footnotes - Listed at end
• Links - Clickable source links

Safety & Content Settings

Content Filtering

Control what content the AI will engage with:

• Standard - Block harmful content
• Strict - Additional filtering
• Custom - Define specific rules

Out-of-Scope Handling

What happens when questions are outside your knowledge base:

Competitor Mentions

How to handle competitor questions:

• Neutral - Don't engage
• Redirect - Focus on your offering
• Compare - Provide comparison (if content exists)

Conversation Settings

Greeting Behavior

Default greeting configuration:

• Automatic - Greet when widget opens
• On First Message - Greet after customer types
• Custom - Define specific trigger

Conversation Timeout

When to consider conversations ended:

• 5 minutes - Quick sessions
• 15 minutes - Standard (recommended)
• 30 minutes - Extended support
• 1 hour - Long-form assistance

Conversation Resume

Allow customers to continue previous conversations:

• Enabled - Show "Continue where you left off?"
• Disabled - Always start fresh

Escalation Defaults

Auto-Escalation Threshold

When AI should automatically offer escalation:

Escalation Prompt

Default message when offering escalation:

Would you like me to connect you with a support specialist 
who can help further?

Customize to match your brand voice.

Learning Settings

Memory Defaults

Default memory behavior for new personas:

• Enabled - Remember customer information
• Disabled - No memory by default

Feedback Collection

Whether to ask for feedback:

• Always - After every conversation
• Sometimes - After resolved conversations
• Never - Don't collect feedback

Saving Changes

After adjusting settings:

1. Review your changes
2. Click Save
3. Changes apply to new conversations
4. Existing conversations unaffected

Persona Overrides

Remember: These are defaults. Individual personas can override:

• Response length
• Escalation behavior
• Language settings
• And more

Configure persona-specific behavior in AI Agents > Personas.

Best Practices

Start Conservative

Begin with conservative settings:

• Standard response length
• Early escalation offers
• Safety filtering enabled

Adjust based on experience.

Test Changes

After changing settings:

1. Use widget preview
2. Test affected scenarios
3. Verify behavior is correct
4. Monitor production impact

Document Rationale

Note why you chose specific settings:

• Helps future decisions
• Useful for team discussions
• Aids troubleshooting

Review Periodically

Revisit settings regularly:

• Monthly at minimum
• After major content changes
• When support patterns shift