Amazon personalize provides two use cases, the first one is recommended products that does not takes customers interaction with products into account, providing recommended products based on other products, and the other takes that interaction data into account and provides recommendation for each specific user

During this project I used Laravel, MySQL, AWS S3, S3 select, PgSQL, Branch.io and developed two apis for products based recommendations and users based personalize recommendations & recommendations were directly being fetched from AWS S3 via S3 select. the models were trained twice each week for the data that was accumulated in that week.

I developed a recommendation system for Grocer Apps which used the amazon personalize service, we uploaded our customers, products and customer-product interaction data for training to this service, It trained certain models on the basis of this data and ultimately give recommendations against products and users which we in return show to our users.