I built around the Zilog80 processor while completing my studies as electronics engineer in 1983, and then joined the research unit in Computer Protocols and Networks of the University of Liège as a research engineer. With the help of an European research programme and industry partners in Belgium, Germany and France, we built an optical Backbone Wideband Network of 140 Mbps in advance of the FDDI standards. Distributed computing and application integration has guided my career since then.

In 1988, I was proposed by the director of a very small firm based in the south of Belgium (STESUD) to lead a software product line for something called EDI; that was about all the requirements! Everything had to be done from designing the product to finding the market, and indeed hiring a team. The entrepreneurship spirit behind the project triggered my decision to accept the challenge. That was the start of an incredible story that drove a little team of Belgian folks to sell software all around the world to some of the most prestigious companies, and making double digit returns by the same token.

We were in 1988, I was in passion for object-oriented programming since I saw a demonstration of Smalltalk and LOOPS at the beginning of the eighties. But Smalltalk was too marginal, Eiffel was not appropriate for the project, and I could only get C programmers. C++ was just getting popular but I never liked it: it didn't solve any of the problems of C - memory allocation notably - and had no dynamic binding at the time. How could you call it an object-oriented language? So we used Objective-C, a true object-oriented extension of C. It was just like Java but with a C syntax, and 10 years earlier!

I was also found of another paradigm smelling like poor practice at the time: scripting... what I really liked in it was the concept of meta-languages and interpreters: no compilation, no floundering in system resources (did I hear you whispering "alike Java"?). Indeed we have nowadays meta-languages for everything, many being descendants of XML. And scripting (JSP, ASP, PHP, VBS, XSLT, ...) is no longer a shame isn't it? So we mixed Objective-C, UNIX shell scripts, C source code for specialized modules, and some of our own meta languages.

We started to develop very specialized products, the only way forward for a very small company with no references. We found a few partners (another story in itself), and developed quite original software capable of handling any format through any protocol. The products were named CODEM - a multi-format translator, EDISC - the clearing system and message broker, EDISAFE - a simpler front end, and TRADEGATE - a windows-based front end. You may still get a look at some of the that I wrote.

I wasn't alone: the director of the small firm attracted the necessary funding from the regional authorities and his assistant - we became friends - was a commercial genius.

We ended up in deploying an EDI Clearing System at SITA (1992) - the most prestigious airline networking company, soon followed by Korean Trade Network, Singapore Telecom, Telepac Portugal, Honk Kong Article Numbering Association, Taiwan Telecom Networks, Embratel - the Brazilian PTT, the European Commission (TRANSIT project) and even the IRS and Customs departments of the US Treasury. We also had successful projects for message brokers and other specialized components with many companies in South Africa, the 28 shipping ports of Spain and the customs authorities, Royal Bank of Canada, Cargolux - world leading all freight airline; just to name a few. Many systems are still in operation, including the one at SITA (upgraded and extended a few times).

In 1997, the company was acquired by Ernst & Young and joined the IT branch of the group named E&Y Consulting. That was the time I really learned what consultancy was all about. I conducted executive missions in Belgium and abroad regarding complex enterprise systems, integration platforms and some portal contexts. I participated in audits of IT departments, strengthening my technical skills with governance and organizational dimensions. The deep value creation and knowledge management culture of Ernst&Young made these years most productive. Sales of our original EDI software continued to scale up, now applied to the distinguished EAI and B2B integration contexts, and with versions significantly extended to support all Internet protocols and formats. A JVM, Java code and JSPs were added to the original software mix.

May 2000 was marked by the fusion between E&Y Consulting and Cap Gemini. It was full of promises but three years later, none of the customer-oriented spirit and knowledge management culture of Ernst & Young was left. My own view is that these have been respectively replaced by profitability objectives and industrialized competency profiles.

That was not my cup of tea. Worse even: we learned that Cap Gemini considered software product development as being incompatible with their vision of an indenpendent software integrator. Sales activities around our original message broker fell to zero and the new J2EE-based version that was in development was stopped just as we generated the first leads in partnership with BEA Europe.

I could have been staying at Cap Gemini indeed, but for a career that would be far from my aspirations. So I decided in December 2002 that it was time to take another route.

If you are a confirmed optimistic, adverse situations always carry something positive: the lack of proper knowledge sources combined with my intense willingness to create value for customers, and the unsatisfied quest of some colleagues for Application and Enterprise Integration expertise altogether forced me to structure concepts and develop my own ideas. Although I started to take distance with the Cap Gemini organization, I was in charge, because of my profile and level in the company, of very interesting customers (complex and demanding ones).

Possibly, my work for an express courier airline marked the turn: the company had to revamp three quarters of its IT infrastructure without stopping operations. Building the proposal was a formidable challenge. Airlines have that particularity that their business is much "segmented": air movement, flight dispatch, capacity planning, ground transportation, shipment, scheduling, weight and balance, crew, maintenance, etc. Mapping such "departments" onto the EAI infrastructure was a natural question: my Business segmentation concept took shape. Since that project, I had the opportunity to test and refine this concept which brought along many others, and notably a series of quite useful integration patterns.

I also had the opportunity to deal with a few of the leading application integration packages (at the time: TIBCO, Seebeyond, IBM Crossworlds, MS Biztalk, Webmethods and BEA Weblogic Integrator): Although they all aim to the same purposes, the architectural differences are staggering. Yet, many problems are left unresolved; no one helps in properly structuring the integration landscape, no one provides a good message store to support the implementation of complex processes and they are too complex for the simple things; few were efficient when it comes to operations. At the same time, me and my team were migrating some of our original software components into an application server, and conducted many brainstorms on the subject matters.

I was also teaching students at the University, having to explain the basics of computer application networking issues from the early EDI days to modern application servers in ...2 hours!

It was a melting pot which made me feel like saturated of concepts, cases, ideas, procedures, contexts...plus many, many questions. I had to do something, I had to offload,
so I started to write.

And when you want to write, the natural step nowadays is possibly to make a... web site. Here we are!

Since October 2003, I'm an independent Enterprise Integration Architect.

Since mid 2004, I have been working a lot with MS Biztalk, became certified developer on SAP XI, made a report on for the European Commission, conducted several audit missions for the largest Energy Company in Belgium, and seek to maintain a high level of pertinence on other platforms like WBI.

[www.artofe.biz] © 2005 bernard Hauzeur • All rights reserved.