Systems and methods for intelligently providing supporting information using machine-learning

A system, comprising:

• one or more processors;
• and a non-transitory computer-readable storage medium containing instructions which, when executed on the one or more processors, cause the one or more processors to perform operations including:
• collecting, at an expense tracking system, a training data set for training a machine-learning model, the training data set representing one or more travel events that previously occurred, each travel event of the one or more travel events corresponding to an expense incurred by a user and one or more event parameters that identify a characteristic of the travel event;
• defining one or more evaluation metrics using the one or more event parameters, each evaluation metric classifying the one or more travel events into an event type;
• training the machine-learning model using the training data set, the trained machine-learning model comprising a plurality of nodes including at least a first node associated with a first event type, the training comprising assigning a weight to each of the plurality of nodes and generating one or more trained correlations between at least two nodes of the plurality of nodes, each node representing a value associated with a travel event of the one or more travel event, wherein each of at least one node of the plurality of nodes corresponds to a hierarchical tree structure;
• detecting a first communication from a computing device operated by a particular user, the first communication corresponding to a request to initiate a process associated with a particular travel event, the request being associated with an actual or anticipated expense, the particular travel event corresponding to the first node having a first event type, and the first communication being received at the expense tracking system;
• in response to detecting the first communication, determining one or more variables from the request, each variable of the one or more variables representing a characteristic of the particular travel event, wherein the one or more variables comprise a client variable that includes a request for a reimbursement for a portion of the actual or anticipated expense;
• in response to determining the one or more variables, accessing one or more rules associated with the client variable, wherein the one or more rules includes selecting nodes representing values associated with a travel events associated with a reimbursement action;
• mapping the one or more variables to the plurality of nodes of the machine-learning model;
• identifying, based at least in part on the mapping, one or more nodes for each of the one or more variables, the one or more nodes being included in the plurality of nodes of the machine-learning model, and the one or more trained correlations being used to identify the one or more nodes, wherein the identified one or more nodes are restricted to a subset of nodes based on the accessed one or more rules, and wherein restricting the identified one or more nodes to the subset of nodes comprises including nodes representing values associated with a travel events associated with a reimbursement action in the subset of nodes;
• retrieving one or more travel event values, each travel event value of the one or more travel event values associated with an action corresponding to an event type;
• generating, by inputting the one or more variables into the trained machine-learning model and based on the one or more travel event values, an output predicting supporting information that identifies a plurality of actions comprising one or more first actions and one or more second actions, each action of the plurality of actions having occurred in association with the particular travel event, and wherein each of the plurality of actions are of a different event type from the particular travel event and at least one action of the plurality of actions is of a second event type;
• and transmitting a second communication to the computing device, the second communication being responsive to the first communication, comprising the one or more travel event values, and identifying the supporting information.