We process an average of 30,000 documents a day. These document images are unstructured. The production flow comprises multiple processes, starting from receiving of document images, OCRing (Optical Character Recognition) to convert image into text; running algorithms (comprising: Python scripts, Regular Expressions, Natural Language Processing, Machine Learning, and .Net programming) to extract the field data elements, validating the extracted data elements against truth tables (in MySQL & Mongo DB). All these processes run automatically on hundreds of auto-systems (Servers).