An AI service that delivers a personalized user experience
Try Cognitive Services for free
Sign-in to Continue
You're almost ready to start building with your 7-day free evaluation.
Sign-in with your preferred account to get started
Increase engagement using personalized interactions
Create rich, personalized experiences for every user of your app. Prioritize relevant content and user experiences, improving app satisfaction, usability, and engagement, with Azure Cognitive Services Personalizer Preview. Only Azure delivers this powerful reinforcement learning–based capability through an easy-to-use API. Reinforcement learning is the set of techniques that allow artificial intelligence (AI) to optimize for a goal based on your configuration.
Personalized experiences in your apps
No machine-learning expertise required
Available on-premises or in Azure
User-friendly interface to easily manage the reinforcement-learning loop
Deliver rich, personalized experiences for every user
Give every user a relevant experience that improves over time. The Personalizer reinforcement-learning cycle learns in real time from your end users’ interactions and helps you improve the user experience to achieve your business goals. Unlike recommendation engines that offer a few recommendations from a large catalog, Personalizer returns its single best recommendation for a user.
Get to market quickly
Get up and running by adding two lines of code. Easily inspect the accuracy of predictions and optimize as needed. Personalizer works with attributes you have about users, content, and products, without data manipulation.
Deploy anywhere, from the cloud to the edge
Personalizer is available where you need it most: in the cloud or as a microservice running locally, where low latency is critical. No matter where your application lives, Personalizer delivers high-throughput training in Azure based on real-time reinforcement signals as they’re processed by your application.
Understand how the Personalizer algorithms work
Personalizer supports model explainability, reproducibility of results, and ongoing evaluation of its effectiveness for added transparency.
Security for the enterprise
Personalizer is offered at a preview rate
Personalizer is in preview and offered at reduced rate. Notice will be provided prior to the end of preview. GA price will be announced in the future when the service becomes generally available.See Personalizer pricing
Get started with Personalizer in three steps
Frequently asked questions about Personalizer
Personalizer is available in preview in West US 2 and West Europe.
There is no service-level agreement (SLA) for previews.
Personalizer learns from the collective behavior of users, through the features or attributes about them that you send to the service. It uses this information to train a shared model that’s updated with information from every interaction, improving personalization outcomes for all users.
Personalizer works best when a rank call has 50 or fewer items. To personalize a choice from a larger list or catalog, use a recommendation engine or sorting technique to reduce the number of items.
Personalizer uses different algorithms for learning and exploration. We’re continuously adding innovations to the system from our work in Microsoft Research and applied use of Personalizer in Microsoft businesses such as Xbox and Windows. See the documentation for a description of the approaches used in Personalizer, and visit the reinforcement learning page to learn what Microsoft Research is doing to advance it.