Learned Style

Teach your agent to communicate in your brand's voice

Learned Style#

Learned Style allows your AI agent to communicate using your brand's unique voice and personality.

What is Learned Style?#

Instead of generic AI responses, Learned Style trains your agent to:

• Match your communication tone
• Use your brand's terminology
• Mirror how you actually talk to customers
• Maintain consistent personality

How It Works#

1. Import Conversations: Upload real conversations from Facebook, Instagram, or email
2. Analysis: AlonChat analyzes your communication patterns
3. Style Generation: A personalized style prompt is created
4. Application: Your agent adopts your communication style

Creating a Learned Style#

From Facebook Messenger#

1. Export your Facebook Messenger data
2. Go to Sources → Chat Training
3. Upload the export file
4. Select conversations to analyze
5. Click Generate Style

From Instagram#

1. Export your Instagram DM data
2. Follow the same process as Facebook

Style Components#

Reviewing Your Style#

After generation:

1. Go to Settings → Learned Style
2. View the generated style profile
3. Review example responses
4. Adjust if needed
5. Activate the style

Multiple Styles#

You can create multiple styles for different purposes:

• Support Style: Empathetic, solution-focused
• Sales Style: Enthusiastic, benefit-focused
• Technical Style: Precise, detailed

Switch between styles in agent settings.

Fine-Tuning#

Manual Adjustments#

Edit the style prompt directly to:

• Add specific phrases to use
• Add phrases to avoid
• Adjust tone levels
• Include brand terminology

Re-training#

If the style isn't quite right:

1. Add more conversation data
2. Click Regenerate Style
3. Review and activate

Best Practices#

1. Use quality conversations: Select your best customer interactions
2. Include variety: Different types of questions and situations
3. Review results: Check sample responses before activating
4. Iterate: Refine based on feedback