A system for extracting and providing information from a source document includes a memory and a processor. The memory stores data associated with the source document. The processor is configured to classify the source document as a determined document type from a plurality of document types. The processor extracts data from the source document based at least in part upon the determined document type and store that extracted data as a first extracted data in the memory. When a request for information is received, the processor then determines if the first extracted data includes all the requested information and if it does not, the processor extracts second extracted data from the image file and stores the second extracted data along with the first extracted data and the image file in the memory. The requested information is then provided to the requesting process from the first and second extracted data.