Eight years ago, I moved to the United States with one goal in mind: building the first open source transactional workflow platform. I had drawn its architecture on a napkin â€" literally, and found some people on Google who could help me put it together. Today, Intalio released Intalio|BPMS 5.0, or what I really consider to be the first working implementation of the original vision. While I will keep the full story of this incredible journey for future articles â€" or a book maybe â€" I would like to share some of its highlights with you tonight.
The original idea came from the intuitive understanding that traditional software development platforms would not scale when applied to the development of business applications built for the Web. It was also motivated by the fact that I am a pretty bad programmer, and that code-less application development environments are a much better option for people like me. Back in 1999, XML had just been invented, J2EE was around the corner, and a fancy way of describing what I had in mind was transactional workflow. Looking for a cost-effective way to build my dream platform, I stumbled upon the nascent open source movement, liked the idea, and mixed all these buzzords together into a pitch that convinced a couple of early stage investors. Before I knew it, we had a team of a dozen engineers laying the foundation for what would be called a Business Process Management System (BPMS) a year later.
Back in the Summer of 2000, there was no such thing as BPM. All the market was interested by was B2B and EAI. Reading through lots of white paper, I pinpointed a common thread â€" processes, dropped the "workflow" moniker in favor of the much cooler "process" â€" which convinced Intalio's co-founder and CTO that the whole idea had some value, and founded a non-profit (BPMI.org) which aim was to develop the set of standards that would govern a new industry. We knew the idea had some merit, and $6M were raised in less than 24 hours with a simple email stating that we would build "webMethods + Vitria in a box." What we did not know was that it would take seven years to put the whole thing in the damn box.
Building enterprise infrastructure software is difficult. Very difficult. And it takes time. A long time. Time is money, so it takes a lot of money too. So much so that most companies trying to build a new piece of enterprise software fail along the way, or get acquired before they get a chance to really make a difference in the market. Things are even more difficult if you're trying to create a new industry, which is pretty much what we were attempting.
Ignorance is bliss, and not knowing what we were up to is what allowed us to keep trying for so long. But timing is everything, and you cannot create a new market category if the market is not ready for it, no matter how hard you try. All you can do is wait for the market to get ready, then run as fast as you can. Of course, we did not know that back then, so we tried to run ahead of the market from 2002 to 2004, with very poor results. Once we finally got it, we decided to wait for the best part of 2005, then started running again in 2006. From February 2006 to September 2007, we went from 12 to 200 customers. The market was now ready for what we had to offer, and we were ready for the opportunity that was coming, finally.
Of course, getting ready for it took time and was painful. Very painful. Lots of money was spent, lots of code had to be re-written, and several friendships were lost along the way, which I very much regret. Nevertheless, through all these years, we fought hard to keep the vision alive, always questioning ourselves, but never doubting that we could make things better. We knew that our platform should be process-centric rather than object-oriented, but we had no control over which standards the industry would rally around. We knew that open source would play a key role in the formation of such a revolutionary platform, but we had a hard time figuring out a working business model around it. And we knew that we needed to build the product's core components first (Designer, Server, and Workflow), but we did not know how to sell it to customers without fancy Business Activity Monitoring.
Eventually, standards were established (BPMN and BPEL), a business model was figured out, and the missing components were added. It took seven years, lots of blood, sweat, and tears, but as far as I am concerned, it was worth every drop of them. Tonight, we have released one of the finest pieces of software engineering I have come across in the 25+ years I have had a love affair with computers, and the last time I felt so proud is when my wife and I gave birth to a cute little snow flake of our own making. For this, I would like to extend my infinite gratitude to all the people who made this possible, from the early dreamers of the first ExoLab Session, to all the fine contributors I have had the privilege of working with at Intalio, and to the customers, partners, and investors who entrusted us with their time and money. To you all, I would like to tell you that your contribution is very much appreciated, and has turned into an amazing piece of software that will have profound impact on the industry for years to come.
Thank you!
Link to original post