Language Support Configuration

Objective:

This documentation is intended to guide the support team in helping users understand, configure, and troubleshoot multilingual capabilities—particularly for less-common languages such as Greek, Romanian, Arabic, and Hebrew. It also expands on our existing documentation by providing detailed setup instructions and practical support workflows.

When to Use This:

• Customers inquire about support for languages not listed under fully supported or beta.

• Users need help configuring agents for multilingual interactions.

• Tickets involve voice clarity or response logic in a secondary language.

Supported Language Overview: https://docs.google.com/spreadsheets/d/1VZKMjH2MHurRwzbWarU5z8qS6iOXVYpLQwqUQqgOt3o/edit?gid=1489477955#gid=1489477955

Step-by-Step: Configuring Multilingual Agents

1. Agent Setup

   • Navigate to your Agent Dashboard.

   • Click the "Voice" tab.

   • Select the desired primary language from the dropdown.

   • For multilingual functionality, choose "Multilingual" if the agent needs to recognize and switch between languages.

2. Voice Model Compatibility

   • Ensure the selected voice engine supports your target language.

   • For unsupported voices, consider switching to multilingual voice mode

3. Prompt Design

   • Write prompts in a neutral tone with appropriate translations.

   • Use short, clear phrases for better voice synthesis accuracy.

   • For RTL languages (Arabic, Hebrew), verify text rendering and spacing.

4. Testing the Agent

   • Use test calls to simulate language-specific scenarios.

   • Adjust prompt wording based on clarity during synthesis.

   • Validate pronunciation and grammar using live input and output examples.

Best Practices:

• Default to fully supported languages unless the customer has a specific use case.

• For beta languages, clearly set expectations about limitations and testing status.