Configure your narration pipeline from end to end: deploy cloud infrastructure, select your workforce identity provider, test open-source voice models, and generate your custom embed script.
Step 1: Deploy Backend
Deploy your private cloud resources
Auritus runs inside your AWS account with no standing third-party access. AWS CDK boots DynamoDB, API Gateway, Cognito User Pool, and the AWS Batch GPU fallback.
Terminal
# Clone and install dependencies
git clone https://github.com/AnthusAI/Auritus.git
cd Auritus
# Deploy backend to your AWS account
auritus deploy --region us-east-1
Step 2: Operator Identity
Choose your authentication method
Select an identity provider. Selecting an option smoothly slides in the specific configuration steps and CLI login command.
Deploy the backend stack with CDK (provisions Cognito User Pool automatically).
Create an operator user in Cognito via AWS CLI or Cognito Console.
Authenticate on your local desk using `auritus login --username <email>`.
Operator Login Command:
auritus login --username <email>
Step 3: Speech Generation
Select your default TTS model
Compare open-source voice models side-by-side. Listen to real speech generated from each model to choose the right balance of speed and expressiveness.