Skip to content

README.md

Latest commit

 

History

History
283 lines (246 loc) · 13.5 KB

README.md

File metadata and controls

283 lines (246 loc) · 13.5 KB

Intelligent Recommendations Accounts

Costa Rica

GitHub GitHub Cloud2BR OSS - Learning Hub

Last updated: 2026-01-29

When creating a new Intelligent Recommendations account: You can create and manage Azure resources for Intelligent Recommendations via the Intelligent Recommendations Azure Resource Manager (ARM) portal. You'll create three resources: Account (Parent resource), Serving (Child resource), and Modeling (Child resource).

Table of Content (Click to expand)

Key Features:

  • Provides personalized product recommendations and telemetry insights using modern machine-learning algorithms.
  • Helps improve catalog navigation and item discovery.
  • Creates upsell and cross-sell opportunities.
  • Improves shoppers' experiences and product usability.
  • Supports various outcomes such as “shop similar looks,” “shop by description,” “real time,” “session based”, Item based recommendations that can combine User interactions and Item Metadata.
  • Can promote and personalize any content type, such as sellable products, consumable media, documents, videos, and more.

Capabilities:

  • Built-in world-class AI-ML delivers personalized results within minutes.
  • Easy to integrate and extend in any ecosystem and experience.
  • Trustworthy software service at scale.
  • Highly adaptable to custom business scenarios and logic, based on input data and algorithm choice.

Note

Here are some use cases:

E-commerce Personalization - Intelligent Recommendations Accounts:

Azure Intelligent Recommendations can be used to provide personalized product recommendations to users based on their browsing and purchasing history. This can significantly improve the user experience and increase sales.

Important

Here’s an example of an end-to-end solution using Azure resources and Azure Intelligent Recommendations Accounts: This solution leverages Azure Intelligent Recommendations to provide personalized product recommendations to users based on their browsing and purchasing history. This can significantly improve the user experience and increase sales.

Architecture: The following Azure services are used in this solution

  • Azure Intelligent Recommendations: For generating personalized product recommendations.
  • Azure Cosmos DB: For storing user browsing and purchasing history.
  • Azure Functions: For processing user data and interacting with Azure Intelligent Recommendations.
  • Azure App Service: For hosting the e-commerce website.

Workflow

  1. User Browsing and Purchasing History: User interactions with the e-commerce website are tracked and stored in Azure Cosmos DB.
  2. Data Processing: An Azure Function is triggered on a regular basis (or in response to changes in the Cosmos DB) to process the user data and generate a batch of recommendation requests.
  3. Generate Recommendations: The batch of recommendation requests is sent to Azure Intelligent Recommendations, which returns a set of personalized product recommendations for each user.
  4. Display Recommendations: The product recommendations are displayed to the user on the e-commerce website hosted on Azure App Service.

Code Samples: Here are some code snippets to illustrate how these components might interact. Note that these are simplified examples and may need to be adapted for a real-world application.

  • Storing User Data in Cosmos DB:
# Python code for storing user data in Cosmos DB
from azure.cosmos import CosmosClient

# Initialize Cosmos Client
cosmos_client = CosmosClient("<Your Cosmos DB Connection String>")

# Select database
database = cosmos_client.get_database_client("<Your Database Name>")

# Select container
container = database.get_container_client("<Your Container Name>")

# Insert user data
user_data = {"id": "1", "browsing_history": [...], "purchase_history": [...]}
container.upsert_item(user_data)
  • Generating Recommendations with Azure Intelligent Recommendations:
# Python code for generating recommendations with Azure Intelligent Recommendations
import requests

# Define the API endpoint
api_endpoint = "https://westus.api.cognitive.microsoft.com/recommendations/v4.0/models/<Your Model Name>/recommend/item"

# Define the headers
headers = {
    "Ocp-Apim-Subscription-Key": "<Your Azure Intelligent Recommendations Key>",
    "Content-Type": "application/json"
}

# Define the payload
payload = {
    "itemIds": ["<Item ID 1>", "<Item ID 2>", ...],
    "numberOfResults": 10,
    "minimalScore": 0.0
}

# Send the POST request
response = requests.post(api_endpoint, headers=headers, json=payload)

# Print the recommendations
print(response.json())
graph LR
    A[User Browsing and Purchasing History]
    B[Azure Cosmos DB]
    C[Azure Functions]
    D[Azure Intelligent Recommendations]
    E[Azure App Service]
    F[Data Processing]
    G[Generate Recommendations]
    H[Display Recommendations]
    A --> B
    B --> C
    C --> D
    D --> E
    B --> F
    F --> G
    G --> H
Loading

Content Discovery - Intelligent Recommendations Accounts:

In media or news platforms, Azure Intelligent Recommendations can suggest articles, videos, or other content that might interest the user, enhancing user engagement and time spent on the platform.

Important

Here’s an example of an end-to-end solution using Azure resources and Azure Intelligent Recommendations Accounts: This solution aims to enhance user engagement and time spent on media or news platforms by suggesting articles, videos, or other content that might interest the user.

Architecture

  1. Data Ingestion: Use Azure Data Factory to ingest data from various sources into Azure.
  2. Data Storage: Store the ingested data in Azure Blob Storage for unstructured data (like videos, images) and Azure Cosmos DB for structured or semi-structured data (like user profiles, article metadata).
  3. Data Processing: Use Azure Databricks to process and transform the data into a suitable format for the recommendation engine.
  4. Recommendation Engine: Use Azure Personalizer, part of Azure Cognitive Services, to generate personalized recommendations based on the processed data.
  5. User Interface: Display the recommended content to the user through a web or mobile application hosted on Azure App Service.

Implementation Steps

  1. Set up Azure Data Factory: Create pipelines to ingest data from various sources into Azure.
  2. Set up Azure Blob Storage and Azure Cosmos DB: Store the ingested data.
  3. Set up Azure Databricks: Write Spark jobs to process and transform the data.
  4. Set up Azure Personalizer: Train a model with the processed data to generate personalized recommendations.
  5. Set up Azure App Service: Develop and deploy your web or mobile application.
graph TB
    A[Azure Data Factory] -->|Ingest Data| B[Azure Blob Storage and Azure Cosmos DB]
    B -->|Store Ingested Data| C[Azure Databricks]
    C -->|Process and Transform Data| D[Azure Personalizer]
    D -->|Generate Personalized Recommendations| E[Azure App Service]
    E -->|Display Recommended Content| F[Users]
    A -->|Set Up Azure Data Factory| G[Implementation Steps]
    B -->|Set Up Azure Blob Storage and Azure Cosmos DB| G
    C -->|Set Up Azure Databricks| G
    D -->|Set Up Azure Personalizer| G
    E -->|Set Up Azure App Service| G
Loading

Ad Targeting - Intelligent Recommendations Accounts:

Advertisers can use Azure Intelligent Recommendations to target users with ads that are relevant to their interests and online behavior, increasing the effectiveness of ad campaigns.

Important

Here’s an example of an end-to-end solution using Azure resources and Azure Intelligent Recommendations Accounts: This solution leverages Azure Intelligent Recommendations to target users with ads that are relevant to their interests and online behavior, thereby increasing the effectiveness of ad campaigns.

Architecture: The following Azure services are used in this solution

  1. Azure Databricks: For data processing and feature extraction.
  2. Azure Machine Learning: For model training and deployment.
  3. Azure Cosmos DB: For storing user profiles and ad data.
  4. Azure Intelligent Recommendations: For generating personalized ad recommendations.
  5. Azure Functions: For real-time ad serving.

Workflow

  1. Data Ingestion and Processing: User behavior data and ad data are ingested into Azure Databricks, where they are processed and features are extracted.

    # Code snippet for data ingestion and processing

  2. Model Training: The processed data is used to train a recommendation model in Azure Machine Learning.

    # Code snippet for model training

  3. Data Storage: The user profiles and ad data are stored in Azure Cosmos DB.

    # Code snippet for data storage

  4. Recommendation Generation: Azure Intelligent Recommendations is used to generate personalized ad recommendations based on the user profiles and ad data.

    # Code snippet for recommendation generation

  5. Ad Serving: Azure Functions is used to serve the recommended ads to the users in real-time.

    # Code snippet for ad serving
graph LR
    A[User Behavior Data] -->|Ingest| B[Azure Databricks]
    C[Ad Data] -->|Ingest| B
    B -->|Process and Extract Features| D[Azure Machine Learning]
    D -->|Train Model| E[Azure Cosmos DB]
    E -->|Store User Profiles and Ad Data| F[Azure Intelligent Recommendations]
    F -->|Generate Personalized Ad Recommendations| G[Azure Functions]
    G -->|Serve Ads in Real-Time| H[Users]
    B -->|Data Processing| I[Code Snippet for Data Ingestion and Processing]
    D -->|Model Training| J[Code Snippet for Model Training]
    E -->|Data Storage| K[Code Snippet for Data Storage]
    F -->|Recommendation Generation| L[Code Snippet for Recommendation Generation]
    G -->|Ad Serving| M[Code Snippet for Ad Serving]
Loading

Learning Path Suggestions - Intelligent Recommendations Accounts:

In educational platforms, Azure Intelligent Recommendations can suggest courses or learning paths to users based on their learning history and career goals, helping them to achieve their learning objectives more efficiently.

Important

Here’s an example of an end-to-end solution using Azure resources and Azure Intelligent Recommendations Accounts: In educational platforms, Azure Intelligent Recommendations can suggest courses or learning paths to users based on their learning history and career goals, helping them to achieve their learning objectives more efficiently.

Architecture: The following Azure services are used in this solution

  • Azure Cosmos DB: Stores user profiles, including learning history and career goals.
  • Azure Databricks: Processes the data from Cosmos DB and prepares it for the recommendation engine.
  • Azure Machine Learning: Trains a model using the processed data to provide course recommendations.
  • Azure Intelligent Recommendations: Uses the trained model to suggest learning paths to users.

Workflow

  1. Data Collection: User learning history and career goals are collected and stored in Azure Cosmos DB.
  2. Data Processing: Azure Databricks processes the data from Cosmos DB. This includes cleaning the data and transforming it into a suitable format for the recommendation engine.
  3. Model Training: Azure Machine Learning trains a recommendation model using the processed data.
  4. Recommendation Generation: Azure Intelligent Recommendations uses the trained model to generate course or learning path suggestions for users.
  5. Presentation of Recommendations: The recommended courses or learning paths are presented to the user through the educational platform's user interface.
graph LR
    A[User Learning History and Career Goals] -->|Data Collection| B[Azure Cosmos DB]
    B -->|Data Processing| C[Azure Databricks]
    C -->|Model Training| D[Azure Machine Learning]
    D -->|Recommendation Generation| E[Azure Intelligent Recommendations]
    E -->|Presentation of Recommendations| F[Educational Platform's User Interface]
    subgraph "Data Collection"
    A -.-> B
    end
    subgraph "Data Processing"
    B -.-> C
    end
    subgraph "Model Training"
    C -.-> D
    end
    subgraph "Recommendation Generation"
    D -.-> E
    end
    subgraph "Presentation of Recommendations"
    E -.-> F
    end
Loading
Total views

Refresh Date: 2026-04-07