Outgoing communication planner utilizing machine learning

A system for controlling sending of messages to retailers that utilizes a business-to-business (B2B) marketplace platform to make product orders, the system comprising:

• a data store that stores the messages that are to be scheduled to be sent to the retailers;
• one or more processors interfaced with the data store and configured to:
• use machine learning to train a communications timing model and to train a retailer interest model based on one or more past messages sent to and received by each retailer of one or more retailers;
• use the communications timing model to determine, for each retailer of two or more of the retailers, a preferred timing for receiving one or more of the messages from the B2B marketplace;
• use the retailer interest model to determine, for each retailer of the two or more of the retailers, a respective level of interest for each of a plurality of different message types that may be sent by the B2B marketplace;
• and
• use a decision engine to schedule and control, for each retailer of two or more of the retailers, based on outputs of the communications timing model and the retailer interest model, when and which of the messages are to be sent to the retailer;
• provide, for a given one of the retailers, a plurality of message queues each of which has a respective message transmission rate limit;
• group similar first type of the messages for the given one of the retailers into a same first one of the plurality of message queues, and group similar second type of the messages for the given one of the retailers into a same second one of the plurality of message queues;
• rank the messages stored within each of the plurality of message queues for the given one of the retailers by calculating respective ranking scores for the messages stored within each of the plurality of message queues;
• and
• for each message queue of the first and the second ones of the plurality of message queues for the given one of the retailers, eliminate one or more of the messages grouped within the message queue that have the lowest ranking scores when there are more messages within the message queue than the respective message transmission rate limit for the message queue;
• for the given one of the retailers, use the decision engine to control when and which of the messages are sent to the given one of the retailers by selecting one or more messages to be sent from each message queue, of the first and the second ones of the plurality of message queues, that were not eliminated from the message queue for the given one of the retailers.