Amazon launches Personalize, a fully managed, AI-based referral service

Providing personalized experiences to customers has become easier thanks to Amazon. The Seattle-based company today announced the general availability of Amazon Personalize, an AWS service facilitating the development of websites, mobile apps, content management systems and e-commerce. -mail marketing offering products, providing personalized search results and customizing funnels on the fly.

It is in some AWS regions to begin, especially in the eastern United States (Ohio), the eastern United States (northern Virginia), the western United States (Oregon), in the Asia-Pacific region (Tokyo), the Asia-Pacific region (Singapore) and the EU (Ireland). Additional locations are on the way.

"We are excited to share with our AWS customers the expertise we have developed over two decades of using machine learning to deliver great Amazon.com experiences," said Vice President of Business Intelligence. machine learning at AWS Swami Sivasubramanian. "Customers are asking for Amazon Personalize and we look forward to seeing how they implement these services to satisfy their own end users. And best of all, it is that these artificial intelligence services, such as Amazon Personalize, do not require any machine learning experience to immediately train, adjust, and deploy models that meet the demand. of their business. "

Custom, announced last year at the re: Invent conference of Amazon, is a fully managed service that trains, adjusts, and deploys custom auto-learning models in the cloud by providing the most up-to-date capabilities. necessary infrastructure and managing tasks such as data processing, feature extraction and algorithm. and optimization and hosting. Customers provide a flow of activity from their applications and their websites (clicks, page views, signups, purchases, etc.), as well as an inventory of items they would like to recommend. (articles, products, videos or music, for example) and possibly demographic information. (like age or geographic location), and they receive the results via an API and only pay for what they use.

Customize existing in-store musical instruments, make recommendations on ingredients and flavors, and design custom style combinations.

"At Subway, the customer experience counts. Thanks to Amazon Personalize, we can quickly provide personalized recommendations for our countless ingredients and flavors, tailored to the unique lifestyles of our busy customers, "said Subway's interim Chief of Information. "Amazon Personalize allows our team … to prepare recommendations without the need for machine learning expertise. We look forward to continuing to work with Amazon Personalize to provide the best experience to our customers who want to eat fresh. We have already successfully tested the use of Amazon Personalize to provide recommendations to customers placing orders from our application, and we are excited to be able to develop custom application notifications soon. "

Personalize is the latest fully managed AWS service to achieve general availability after Textract, which uses machine learning to analyze data tables, forms, and entire pages of text and data. In March, Amazon launched AWS Deep Learning Containers, a library of preinstalled Docker images with popular in-depth learning frameworks.