Google’s new service, which uses artificial intelligence to recommend products to online shoppers, is now available to everyone in beta. AI draws heavily on Google’s recommendation technology on youtube and search services. The tool is designed to help retailers analyze customers’ purchase history in order to provide personalized product suggestions. This means that it focuses on individual customers, not products.

retailers can start deploying the tool by importing directory and user event data when creating Google cloud projects. They can then choose a model type, set their own optimization goals, and start training the model. The training process will last two to five days before recommendations can be delivered to customers.

in addition to the beta version, Google has announced a new pricing structure for the tool, which includes three forecast price levels based on capacity and separate costs for model training and adjustment. The company will also provide $600 in credit to new customers who sign up to use the service. In addition, new Google cloud users can also get $300 free points.