Deduction claim document parsing engine

A method for processing documents, the method comprising:

• extracting document information from one or more deduction claim documents by a data extraction module, wherein the document information comprises at least one of tabular data or non-tabular data;
• classifying the one or more deduction claim documents or the document information using a machine learning model, the machine learning model being trained using a plurality of document formats;
• generating one or more data objects associated with at least one corresponding deduction claim based on the classification of the document information using the machine learning model,
• wherein generating the one or more data objects associated with the deduction claim based on the tabular data comprises:
• processing the tabular data with two or more tabular data extraction modules, wherein the two or more tabular data extraction modules process the tabular data in parallel;
• and
• generating the one or more data objects associated with the deduction claim based on the tabular data processed by the two or more tabular data extraction modules;
• wherein generating the one or more data objects associated with the deduction claim based on the non-tabular data comprises:
• processing the non-tabular data with two or more key-value pair data extraction modules, wherein the two or more key-value pair data extraction modules process the non-tabular data in parallel;
• and
• generating the one or more data objects associated with the deduction claim based on the non-tabular data processed by the two or more key-value pair data extraction modules;
• providing an output data file comprising the one or more data objects;
• providing an accuracy assessment for the at least one corresponding deduction claim by comparing the output data file to reference data or ground truth data;
• and
• modifying the machine learning model using at least the accuracy assessment.

A method for processing documents, the method comprising:

• accessing one or more deduction claims;
• extracting document information from the one or more deduction claims, wherein the document information comprises at least one of tabular data or non-tabular data;
• classifying the document information using a first machine learning model, the first machine learning model being trained using a plurality of document formats;
• generating one or more data objects associated with at least one corresponding deduction claim and based on the classification of the document information using the first machine learning model,
• wherein generating the one or more data objects associated with the deduction claim based on the tabular data comprises:
• processing the tabular data with two or more tabular data extraction modules, wherein the two or more tabular data extraction modules process the tabular data in parallel;
• and
• generating the one or more data objects associated with the deduction claim based on the tabular data processed by the two or more tabular data extraction modules;
• wherein generating the one or more data objects associated with the deduction claim based on the non-tabular data comprises:
• processing the non-tabular data with two or more key-value pair data extraction modules, wherein the two or more key-value pair data extraction modules process the non-tabular data in parallel;
• and
• generating the one or more data objects associated with the deduction claim based on the non-tabular data processed by the two or more key-value pair data extraction modules;
• associating the one or more data objects using a second machine learning model with one or more attributes of an output data file;
• and
• generating the output data file based on the second machine learning module comprising the one or more data objects associated with the one or more attributes of the output data file.