EDI: building a business community

The last time that I used the title of this article was for a report that I wrote for the Financial Times, which was published the best part of a decade ago. The premise behind that report was that EDI (Electronic Data Interchange) offered the promise of establishing closer relationships between different organisations that would be able to partner in a variety of new ways. While some progress was made in that direction, the truth is that the proprietary networks that were (and are) involved were too complex to maintain and too expensive. As a result, many organisations did not participate or did so only partially, or only under duress, so that EDI has never really fulfilled its potential.

However, the advent of the Internet on the one hand and XML on the other, together with secure communications protocols and guaranteed deliveries, means that we could be on the verge of a new generation of EDI that can actually start to deliver on its promise.

From my perspective what is interesting about this market is (and this is typical of emerging markets) the number of directions from which vendors are moving into this space. The obvious ones are products like DataStage TX (the ex-Mercator product now owned by IBM), SeeBeyond (now part of Sun), Informatica, and so on, which variously come from the application (and BPM) or data integration space. However, an interesting entrant into the field is Xenos, which derives from somewhere else entirely (and I don’t mean Canada where the company is based).

Xenos has been around for a long time (since the 80s—it went public in 1999) and its historic speciality is in the processing of very high volume printing on the mainframe and, more recently, in the transformation of that data so that it can be presented on the web. The company has also acquired significant technologies from other vendors so that it can handle structured data (as well as things like COBOL copybooks) and semi-structured data (XML) in addition to print stream data.

From a technical perspective the company wins out because it has much higher performance than other solutions. There are a variety of reasons for this: its I/O streaming, its caching and its indexing technology being notable examples. And it is worth noting that the company has just announced a partnership with Oracle (in the UK) to address the financial market.

However, of interest with respect to this article is the company’s terminalONE product.

terminalONE consists of three elements: performance transformation (handling SWIFT messages, HIPAA and so forth as well as EDI); transportation (via EBXML—the company claims the largest such implementation in the world with the Norwegian Government); and traffic routing (with workflow, introspection of messages, guaranteed delivery, security and so on). According to Xenos (and I would not disagree), it is these three issues that drive the complexity and expense of traditional approaches to EDI.

As I have noted there are various vendors converging on the space for extracting structured and unstructured data from front-end sources, transforming it and then loading it, not into a data warehouse but into a receiving application. This raises two questions: first, what do we call this—ETL would be counter intuitive—one vender has suggested data transfer to me? More importantly, can Xenos compete with traditional vendors in this space? The answer to this sort of question usually depends on marketing: Xenos surely has the technology, it remains to be seen whether its marketing can establish a reputation to match.