Content personalization based on reinforcement learning

A system comprising:

• control circutory configured to:
• cause a user device to display a user interface comprising a plurality of carousels, wherein each carousel of the plurality of carousels corresponds to a respective content genre;
• input/output circutory configured to:
• receive, from a server, configuration data that defines:
• (a) a first event of a first event type for a first carousel of the plurality of carousels;
• and
• (b) a second event of a second event type for a second carousel of the plurality of carousels, wherein the second event type is different from the first event type;
• wherein the control circutory configured to:
• monitor based on the configuration data, inputs to the user interface to detect occurrence of (a) the first event defined by the configuration data, and (b) the second event defined by the configuration data
• based on the detecting the occurrence of the first event, transmit to the server a first request to update the first carousel
• receive first update data that was transmitted by the server based on the first request
• based on receiving the first update data, update the first carousel without updating carousels of the plurality of carousels other than the first carousel of the plurality of carousels;
• and
• based on the detecting the occurrence of the second event, transmit to the server a second request to update the second carousel
• receive second update data that was transmitted by the server based on the second request, and
• based on receiving the second update data, update the second carousel without updating carousels of the plurality of carousels other than the second carousel of the plurality of carousels.