Systems and methods for batch multiplexing a plurality of transactions are provided. A system may include a machine-learning (“ML”) based engine configured to receive instructions to execute a first transaction, derive a plurality of sets of delay parameters for the first transaction, and display the sets of delay parameters on a graphical user interface (“GUI”) associated with a sender of the instructions. The engine may be configured to receive a selection of a set of delay parameters, and, based on the selection, append the first transaction to one of a plurality of pending transaction queues. The engine may be further configured to, within a maximum time window, combine the transactions in the pending transaction queue to create a single multiplexed transaction, and execute the multiplexed transaction. The engine may also be configured to demultiplex the multiplexed transaction after it is executed.