Qlytics Document Intelligence leverages Natural Language Processing (NLP) and Machine Learning (ML) to extract information from PDF files. The system first translates PDF files into text, then applies a custom-made text sequence classifier to tag sections, paragraphs, and tables. It then saves the representation to support information retrieval. For retrieval, I designed a scripting language named "Document Query Language" (DQL) that has a similar syntax to Python. With just several lines of code, you get to loop through paragraphs and tables, locate key/value pairs and save the output to a CSV. The entire app is designed like an IDE for documents. You develop DQL scripts in it. Then you run a batch job for similar PDFs against the same script. Extract results can also be piped through some API.
