Automating Deployment of Azure Speech-to-Text Model Across Subscriptions

Question

I have successfully deployed an Azure Speech-to-Text model in my subscription. Now, I need to automate the process of deploying the same model to my colleague's subscription.

 

Current Setup:

I have an existing Speech-to-Text service deployed in my subscription.

The goal is to replicate this deployment to another subscription automatically.

This should include configuration, custom models, and API keys.

Questions:

What is the best way to automate this deployment?

Can I use ARM Templates, Bicep, Terraform, or Azure CLI?

Is there a recommended approach to export and redeploy the model?

How can I transfer the custom Speech-to-Text model?

Is there a way to export/import the trained model from one subscription to another?

Can I use Azure DevOps or GitHub Actions to automate the process?

Any example workflows or pipelines would be helpful.

Additional Context:

Both subscriptions are within the same Azure AD tenant but belong to different users.

I want to minimize manual steps and ensure a fully automated deployment.

Any guidance, best practices, or references to Microsoft documentation would be greatly appreciated!

 

Answer 1

Hi Dinnemidi Ananda Kumar,
Greetings & Welcome to the Microsoft Q&A forum! Thank you for sharing your query.

ARM templates these are JSON files that define the infrastructure and configuration for your Azure resources. You can export the current configuration of your Speech-to-Text service and redeploy it in another subscription.

Bicep is a domain-specific language (DSL) for deploying Azure resources, which simplifies the syntax of ARM templates.

Terraform is an open-source infrastructure as code (IaC) tool that allows you to define and provision infrastructure using a high-level configuration language. It supports Azure and can be used to automate the deployment of your Speech-to-Text service.

Azure CLI can be used to script the deployment process. You can write scripts to create and configure the Speech-to-Text service in the target subscription.

Exporting the Model:
You can use the Azure portal or Azure CLI to export the configuration of your Speech-to-Text service. This includes the custom models and API keys.

Redeploying the Model:

Use the exported configuration to redeploy the service in the target subscription. ARM templates, Bicep, or Terraform can be used for this purpose.

Export/Import Trained Model:

Currently, there isn't a direct way to export and import custom models between subscriptions. However, you can retrain the model in the target subscription using the same data and configuration.

Azure DevOps:

Azure DevOps provides pipelines that can automate the deployment process. You can create a pipeline that uses ARM templates, Bicep, or Terraform to deploy the Speech-to-Text service.

GitHub Actions:

GitHub Actions can also be used to automate the deployment. There are templates available that support the deployment of Azure resources, including custom Speech models.

I hope this information helps.