Azure AI & Machine Learning: Cognitive Services Deployment Samples

Deployment Strategies

Key Concepts

• Managed Online Endpoints: Scalable, real-time inference for your models.
• Batch Endpoints: Process large volumes of data asynchronously.
• Containers: Deploy Cognitive Services on-premises or on Azure Arc.
• CI/CD Integration: Automate your deployment pipelines.

Sample Scenarios

• Deploying Text Analytics as a Container

  Learn how to containerize and deploy the Text Analytics service for local or hybrid cloud environments.

  Containers Text Analytics Docker
  View Details

• Real-time Face Detection Endpoint

  Set up a managed online endpoint for real-time face detection using the Computer Vision API.

  Managed Endpoint Computer Vision REST API
  View Details

• Batch Speech-to-Text Deployment

  Implement a batch inference pipeline for converting large audio files to text using Azure Speech Service.

  Batch Endpoint Speech Service Python SDK
  View Details

• Deploying Language Understanding (LUIS) to Kubernetes

  A guide to deploying LUIS applications to a Kubernetes cluster for flexible scaling and management.

  Kubernetes LUIS DevOps
  View Details

Example Configuration Snippet

Here's a common snippet you might find in a deployment configuration using Azure CLI:

az cognitiveservices account deployment create \
    --name my-text-analytics-deployment \
    --resource-group my-resource-group \
    --account-name my-cognitiveservices-account \
    --kind "TextAnalytics" \
    --sku "F0" \
    --location "eastus" # Or your preferred region

For more complex deployments, consider using ARM templates or Terraform. See the individual sample documentation for specific manifests and scripts.

Getting Started

To start deploying Cognitive Services, you'll typically need:

1. An Azure subscription.
2. An Azure Cognitive Services resource created in your subscription.
3. Appropriate SDKs or CLI tools installed (e.g., Azure CLI, Python SDK).
4. For container deployments, Docker and a container registry.

Refer to the official Azure Cognitive Services documentation for foundational information.