Events
Mar 17, 9 PM - Mar 21, 10 AM
Join the meetup series to build scalable AI solutions based on real-world use cases with fellow developers and experts.
Register nowWhat's new in Personalizer
Important
Starting on the 20th of September, 2023 you won’t be able to create new Personalizer resources. The Personalizer service is being retired on the 1st of October, 2026.
Learn what's new in Azure AI Personalizer. These items may include release notes, videos, blog posts, and other types of information. Bookmark this page to keep up-to-date with the service.
- Personalizer Inference Explainabiltiy is now available as a Public Preview. Enabling inference explainability returns feature scores on every Rank API call, providing insight into how influential each feature is to the actions chosen by your Personalizer model. Learn more about Inference Explainability.
- Personalizer SDK now available in Java and JavaScript.
- Local inference SDK (Preview): Personalizer now supports near-realtime (sub-10ms) inference without the need to wait for network API calls. Your Personalizer models can be used locally for lightning fast Rank calls using the C# SDK (Preview), empowering your applications to personalize quickly and efficiently. Your model continues to train in Azure while your local model is seamlessly updated.
- Auto-Optimize (Preview) : You can configure a Personalizer loop that you are using to continuously improve over time with less work. Personalizer will automatically run offline evaluations, discover better machine learning settings, and apply them. To learn more, see Personalizer Auto-Optimize (Preview).
- Multi-slot personalization (Preview): If you have tiled layouts, carousels, and/or sidebars, it is easier to use Personalizer for each place that you recommend products or content on the same page. Personalizer now can now take a list of slots in the Rank API, assign an action to each, and learn from reward scores you send for each slot. To learn more, see Multi-slot personalization (Preview).
- Personalizer is now available in more regions.
- Updated code samples (GitHub) and documentation. Use the links below to view updated samples:
- New tutorial - using Personalizer in a chat bot
- New tutorial - using Personalizer in a web app
The following are available in Public Preview:
- Apprentice mode as a learning behavior.
- TLS 1.2 is now enforced for all HTTP requests to this service. For more information, see Azure AI services security.
- Personalizer is generally available (GA)
- Azure Notebooks tutorial with entire lifecycle
The following preview features were released at the Build 2019 Conference:
Additional resources
Training
Module
Make recommendations with Azure AI Personalizer - Training
In this module, learn how to use Azure AI Personalizer to make recommendations.
Certification
Microsoft Certified: Azure AI Engineer Associate - Certifications
Design and implement an Azure AI solution using Azure AI services, Azure AI Search, and Azure Open AI.
Documentation
-
Create Personalizer resource - Azure AI services
In this article, learn how to create a personalizer resource in the Azure portal for each feedback loop.
-
Where and how to use - Personalizer - Azure AI services
Personalizer can be applied in any situation where your application can select the right item, action, or product to display - in order to make the experience better, achieve better business results, or improve productivity.
-
How Personalizer Works - Personalizer - Azure AI services
The Personalizer _loop_ uses machine learning to build the model that predicts the top action for your content. The model is trained exclusively on your data that you sent to it with the Rank and Reward calls.