Graphical user interface and profile configuration system for predicting attributes for controlling reconciliation actions

A computer-implemented method performed by a computing system, the method comprising:

• generating and displaying a profile configuration dashboard including a graphical user interface including interactive input fields for defining one or more target profile attributes for prediction in an account profile;
• wherein the account profile is a data structure that defines at least a plurality of profile attributes and criteria for controlling how transactions associated with account combinations are handled during a reconciliation process;
• for a first target attribute defined in the graphical user interface, executing an automated machine learning process including creating and training a plurality of candidate machine learning models from an available set of machine learning models;
• wherein the training includes training each candidate machine learning model to generate a predicted value for the first target attribute using historical account profiles that contain different combinations of profile attributes associated with different account combinations;
• selecting a best-fit machine learning model, from the plurality of candidate machine learning models trained, that generates the predicted value for the first target attribute with an accuracy above a threshold;
• executing the best-fit machine learning model to generate the predicted value for the first target attribute and storing the predicted value in the account profile as one of the plurality of profile attributes and criteria;
• executing query logic to search a database of transactions based on the plurality of profile attributes and criteria in the account profile that match a set of transactions;
• and
• routing and transferring the matching set of transactions into a reconciliation data table assigned to the account profile.

A non-transitory computer readable medium including instructions stored thereon that when executed by at least one processor of a computing system, cause the computing system to:

• generate and display a profile configuration dashboard including a graphical user interface configured to provide input fields for defining one or more target profile attributes for prediction in an account profile;
• wherein the account profile is a data structure that defines at least a plurality of profile attributes and criteria for controlling how transactions associated with account combinations are handled during a reconciliation process;
• for a first target attribute defined in the graphical user interface, execute an automated machine learning process that creates and trains a plurality of machine learning models from an available set of machine learning models;
• wherein the automated machine learning process trains each machine learning model to generate a predicted value for the first target attribute using historical account profiles that contain different combinations of profile attributes associated with different account combinations;
• select a best-fit machine learning model, from the plurality of machine learning models trained, that generates the predicted value for the first target attribute with an accuracy above a threshold;
• execute the best-fit machine learning model to generate the predicted value for the first target attribute and store the predicted value in the account profile as one of the plurality of profile attributes and criteria;
• execute query logic to search a database of transactions based on the plurality of profile attributes and criteria in the account profile that match a set of transactions;
• and
• route and transfer the matching set of transactions into a reconciliation data table assigned to the account profile.