Data discrepancy analysis using machine learning models

A system, comprising:

• a processor;
• and
• a computer readable storage medium having computer readable program code that when executed by the processor performs operations, the operations comprising:
• receiving data from a plurality of source data systems;
• analyzing the data to identify a data discrepancy of a data field;
• generating a recommendation that indicates whether the data discrepancy is fixable;
• in response to the recommendation indicating that the data discrepancy is fixable, fixing the data discrepancy by modifying a value of the data field;
• and
• routing the data to a first downstream system service that does not rely on the data field and to a second downstream system service that does rely on the data field;
• and
• in response to the recommendation indicating that the data discrepancy is not fixable,
• routing the data to the first downstream system service that does not rely on the data field;
• and
• disabling the second downstream system service that does rely on the data field.

A computer-implemented method comprising operations for:

• receiving data from a plurality of source data systems;
• analyzing the data to identify a data discrepancy of a data field;
• generating a recommendation that indicates whether the data discrepancy is fixable;
• in response to the recommendation indicating that the data discrepancy is fixable,
• fixing the data discrepancy by modifying a value of the data field;
• and
• routing the data to a first downstream system service that does not rely on the data field and to a second downstream system service that does rely on the data field;
• and
• in response to the recommendation indicating that the data discrepancy is not fixable,
• routing the data to the first downstream system service that does not rely on the data field;
• and
• disabling the second downstream system service that does rely on the data field.