This is one of my favorite ideas.
I conceived of this component and pitched it to United Guaranty after noticing that every origination system created by UG had to produce very similar printed outputs (mostly commitment certificates). In each case, the systems used their own vastly dissimilar methods of creating and delivering these reports. In one case an AS/400 based faxing system fed by RPG programs is used, in another a combination of Microsoft Word documents and Adobe Acrobat are used to generate PDF files for web delivery (extremely slow and inefficient!). Sometimes HTML delivery is used. In every case, however, the same document format must be delivered to the customer, and it must conform to both industry standards and the standards of the legal department. The proliferation of all these document generation techniques made making changes to a single document a time-consuming and error-prone task.
My idea was to replace these dissimilar generation systems with a single document generation and delivery system. This differs from stock offerings in its flexibility and scalability. Using my design it would be possible to take a document request from any system, regardless of implementation; generate the requested document in any of a number of formats (DHTML, PDF, TIFF); and send it anywhere in the world by any of a number of delivery methods: HTTP, email, fax, FTP, printer, or file. As good as they are, no single commercial offering delivers all of the features of the proposed design with the solidity and performance required for a true enterprise performer. My design combines the best of breed to create an exceptional delivery system.
To accomplish this I proposed a component consisting of a socket listener that feeds a controller. The controller then calls a subsystem to generate the document itself, and calls a delivery subsystem to send the document to the intended recipient. The delivery subsystem would consist of a consistent interface behind which are a number of delivery components, each handling a different delivery method.
In addition to solving the problems mentioned earlier, this design has the added advantage of allowing UG to replace their aging fax delivery system, as there would be an abstract interface to the delivery subsystem. Thus UG could simply drop in a robust alternative like that provides modern load balancing and distribution capabilities. Similarly, any email system could be plugged in, from Lotus Notes to Sendmail.
I took this project from conception through the Functional System Design. I evaluated components for the various subsystems, recommending Actuate for the document generator, RightFAX for faxing, and Sendmail for email delivery. In the process of performing the cost/benefit analysis I noted that the benefits were significant, arising largely from reducing the number of generation systems to maintain, and from the ability to distribute fax delivery geographically to reduce long-distance charges. Other benefits include immediate turnaround for document changes (instead of the pre-existing several week delay) and the ability to give the business unit direct control over the published documents' formats.
I have also created a proof-of-concept using totally cost-free Linux components. The UG project has been prioritized and assigned to a team for implementation and is now awaiting funding.