Intelligent routing control

A system that classifies insurance documents having insurance data, the system comprising:

• an assessment queue storing a plurality of documents in a memory of a device;
• each of the plurality of documents comprise pages containing insurance data, each document having an associated pre-defined destination and associated metadata;
• a plurality of primary nodes representing a plurality of independent processing units that automatically:
• convert a set of documents encoded in a first file format for storage in a non-transitory computer media to a second file format that removes all metadata associated with each document;
• and
• partition each document of the set of documents into separate stand-alone documents such that each partitioned document represents no more than a predetermined physical page length;
• a plurality of secondary nodes representing a plurality of independent processing units that automatically:
• convert each of the partitioned documents into separate recognition vectors that represent information contained in the partitioned documents;
• classify the partitioned documents through a machine learning algorithm comprising a plurality of learning models that are combined to generate a summed output;
• the plurality of learning models includes a successive learning model that minimizes a plurality of residuals generated from a preceding learning model;
• processing the summed output as an input to a subsequent learning model separate from the plurality of learning models that embeds routing data in second metadata associated with each of the partitioned documents to reduce an error;
• merge the classified partitioned documents based on a plurality of rules based at least in part on the second metadata;
• and
• cause the merged documents to be routed to a remote destination independent of a predefined destination.