Whatever Cloud providers you use (e.g., Azure or Amazon), document retrieval can be cumbersome if the search returns more documents than you are expecting. This can result from full text indexing or inappropriate metadata to filter the search. Business documents are easy to classify per category, with relevant references such as customer ID, invoice number or date, etc. For better results, develop a set of metadata to pass to the Cloud application along with the PDF document. Metadata should be defined by reading the document content and selecting the right words, or retrieved from an external resource (such as an email address of the contact point).
Feeding a document cloud storage for easy retrieval requires the ability to identify and pass relevant metadata for each related document.