Project Setup: Advanced Settings

Advanced Settings in Rapport

The Advanced Settings panel in Rapport provides fine-grained control over your project’s behaviour and integration. While many users can launch effective experiences using the main project setup interface, the Advanced Settings menu is available for those who want to customise AI behaviour, manage API credentials, and directly configure TTS/STT providers.

Advanced Settings can be accessed in two ways:

• As the final item on the left-hand menu

• Within other configuration menus (such as AI, Voice)

What You Can Configure in Advanced Settings

1. AI Type & Configuration

Choose from multiple types of AI integrations:

Generative AI

These include providers like OpenAI, Gemini, Groq, and others.

• Credentials: Use Rapport’s built-in credentials or enter your own API key.

• Model: Choose a model from a dropdown list. You can also input a model name manually if it's not listed.

• Prompt (Personality): Define a short prompt to shape how your AI responds.

• Advanced Parameters: Configure temperature, top-p, frequency penalty, and other provider-specific values.

Use this section if you want full control over how your character generates language or interacts.

Assistant AI

Useful if integrating with an external assistant (e.g., GPT Assistant):

• Credentials: Toggle between Rapport’s or your own.

• Assistant ID: Provide an ID linked to the assistant you’ve set up with your provider.

Chatbot AI

For chatbot platforms like Dialogflow, Lex, or Cleverbot:

• Credentials: Choose whether to use Rapport or your own.

• Additional Fields: Vary by provider and are shown contextually.

Other AI Types

Rapport also supports several specialized AI types for advanced or experimental use:

• Echo AI: Repeats the user’s spoken input using the selected voice. Great for testing voices and setups.

• Voice Echo: Uses the user’s own voice to repeat their spoken words.

• Idle (Custom AI): Provides full programmatic control over character behaviour. Requires advanced configuration. Read more

2. Speech-to-Text (STT) Configuration

• Enable/Disable: Toggle speech recognition on or off for your experience.

• Provider Selection: By default, Rapport auto-selects a provider based on language (priority order: Whisper > Azure > Google > AWS)

• Credentials: Use Rapport credentials or uncheck to use your own.

• Language Matching: If using your own STT, only supported languages for that provider will be shown. The provider will not auto-change if the language is changed.

3. Text-to-Speech (TTS) Configuration

• Enable/Disable: Control whether the voice engine speaks responses.

• Provider Selection: Choose between Azure, AWS Polly, Google, or Eleven Labs.

• Credentials: Toggle between Rapport or your own.

• Custom Voice ID: For providers like Eleven Labs, entering a valid voice ID lets you use a custom voice.

When using your own credentials, only voices supported by your provider will be listed. Rapport will not auto-switch providers if the language is changed.

Best Practices

• Use Rapport credentials if you want to get started fast.

• Use your own credentials if you need access to a specific model, voice, or if you're managing usage through your own provider account.

• Test changes to confirm correct voice output or AI behaviour.

Notes on Multilingual Voices

• Multilingual voices now include metadata that limits their availability to supported languages only.

• When a language is selected, only voices that natively or officially support that language will appear.

Summary Table

For more details, visit the specific sections linked from each left-hand menu item.