AWS connect documentation change
Summary
Removed documentation about configuring Nova Sonic-only and beta voices using session attributes, including troubleshooting steps and runtime behavior details.
Security assessment
The change removes a feature configuration section but contains no evidence of security vulnerability fixes or security-related context. It appears to be routine documentation cleanup.
Diff
diff --git a/connect/latest/adminguide/nova-sonic-speech-to-speech.md b/connect/latest/adminguide/nova-sonic-speech-to-speech.md index d76809206..5b59a5cc3 100644 --- a//connect/latest/adminguide/nova-sonic-speech-to-speech.md +++ b//connect/latest/adminguide/nova-sonic-speech-to-speech.md @@ -7 +7 @@ -Part 1: Configure Speech-to-Speech for a Bot LocalePart 2: Configure a Nova Sonic Compatible Voice in a FlowPart 3: Configure Nova Sonic–Only and Beta Voices Using Session Attributes +Part 1: Configure Speech-to-Speech for a Bot LocalePart 2: Configure a Nova Sonic Compatible Voice in a Flow @@ -122,60 +121,0 @@ Choose **Save** , then **Publish** to activate the configuration. -## Part 3: Configure Nova Sonic–Only and Beta Voices Using Session Attributes - -You can configure Nova Sonic–only voices and beta voices for your Amazon Connect Conversational AI bot. These voices are supported by Nova Sonic Speech-to-Speech (S2S) but are not available as Amazon Polly TTS voices. Because they are not surfaced in the Set voice block, you must configure them using session attributes in the Get Customer Input (GCI) block. - -When you use a Nova Sonic–only or beta voice, Amazon Connect will generate Nova Sonic expressive audio for that segment. However, any prompts rendered using Amazon Polly in the same flow may sound inconsistent, because Polly does not support these voices. Choose this option only if you are comfortable with this variability. - -### Step 1: Add a GCI block to your flow - -Open the flow that interacts with your Nova Sonic bot. In the block library, search for **GCI** and drag the block into the designer. - - - -### Step 2: Select Amazon Lex in the GCI block configuration - -In the GCI block configuration panel, select **Amazon Lex**. - - - -### Step 3: Set the bot and alias - -Choose the bot and alias that is configured to use Speech-to-Speech: Amazon Nova Sonic. - - - -### Step 4: Add a session attribute override - -In the Lex Configuration, scroll to **Session attributes** and choose **Add an attribute**. - - - -### Step 5: Set the override key and beta voice name - -Use the following key to override the default S2S voice: **x-amz-lex:audio:speaker-model-voice-override** - -For the value, enter the Nova Sonic–only or beta voice ID. The value must match the exact case expected by the model. For example, the following sets the voice to **tiffany**. - -After adding the key and value, choose **Confirm**. - - - -### Runtime behavior - - * Nova Sonic uses the Nova Sonic–only or beta voice for the duration of the session. - - * If the session later reaches a Set voice block using Amazon Polly, the audio may switch back to the Polly voice. - - - - -### Troubleshooting - - * **Voice not applied:** Confirm that the voice ID is typed exactly as required (case sensitive). - - * **Unexpected voice switching:** Ensure no downstream Set voice block uses a Polly voice. - - * **No S2S response:** Verify that the locale is configured for Speech-to-Speech: Amazon Nova Sonic. - - - -