Welsh wizardry

Written By:
Published:
Content Copyright © 2007 Bloor. All Rights Reserved.

Last year I wrote about the first Welsh IT company that I had come across and now I have discovered another. However, while ETL Solutions was (and is) working in a well-established space, Informavores, the subject of today’s article, is doing something entirely new or, at least doing it in a new way.

It is difficult to describe what Informavores does because it does not fit into any established market niche but, rather, crosses several. Its products provide a model-driven development environment (sort of), a rules engine (but not like any rules engine you’ve previously seen), and a knowledge management tool.

Let me start at the beginning. Regardless of which Informavores product you use (which I’ll come back to) you start by building a decision tree using a conventional graphical approach. The decision tree may be either iconised or use text boxes as in a flow diagram, depending on the product. And that’s it.

Well, not quite. But the logic of the application you are building is now complete and the software will automatically generate the application, along with appropriate web forms, directly from the decision tree (which is known as a “spark”). Just as in a fully model-driven environment the model IS the application so, here, a spark IS the application. You will of course want to test the application and there are automated testing tools provided as well as consistency checking.

Informavores reckons (and it doesn’t seem unlikely) that building applications through the use of sparks is an order of magnitude faster than when using conventional development techniques. Moreover, it is much, much simpler than using standard rules engines. Business Analysts with no programming skills really can build sparks.

The Informavores products are not suited to all types of applications but they are particularly useful when form filling is required (either directly or indirectly) or where complex rules are involved or when you want to test large volumes of data against particular rules.

The first product is called Firebox and it is used for things such as online quotations. Since a number of Informavores customers are in the financial services space, and since the FSA mandates relevant standards for these companies, Firebox includes a complete audit trail, with timestamps, for this sort of environment. There are also query facilities so that you can see the most common paths people took through the process and where they fell off.

The second product is called Firefly and this is used for help desks, contact centres, telesales, procurement and so forth where scripted guidance through the process is required. It is with this product that you will typically use a spark that is more akin to a flow diagram than an iconised decision tree. It is also with this product that the company’s knowledge management capabilities come into their own, capturing the relevant text for each stage of the operation from relevant experts.

Thirdly, there is Firestorm, where you want to check a (large) dataset against one or more rules. Examples of usage in this area might be VAT audits or in ETL (extract, transform and load) environments where you could develop rules that defined which data you wanted to extract.

Despite its relative youth (the company has only been active for a couple of years—it was founded in 2003) it has some big name customers both in the UK and the United States. I am not surprised: Informavores does some cool stuff and is definitely worth checking out.