Barry Klawans gave an excellent talk to the SDForum Business Intelligence SIG when he spoke on "Two Disruptive Trends, Open Source and SaaS Meet Business Intelligence" at the February meting. Open Source is something that I have written about before. SaaS stands for Software as a Service, the idea that information technology can be delivered as a service over the internet. The best example of a successful SaaS business is Salesforce.com. Barry knows the territory well as he is CTO of JasperSoft, an Open Source BI reporting company.
Barry started with the Innovators Dilemma, a book from the 90s that describes how established technologies and products markets can be overturned by innovators who use new disruptive technology or business models. Existing BI software vendors tend to target the high end of the market and they are vulnerable to disruption from new vendors that start by targeting the under served lower end. Barry believes that Open Source and SaaS are the forces that will overthrow the old guard of established BI vendors.
Next, Barry took us through the BI stack and Open Source projects that address it. Successful Open Source projects concentrate on doing one thing and doing it well. One of the problems with using Open Source is that you have to integrate several Open Source packages to build a system. Integration is made more difficult because active Open Source projects tend to have a very short release cycle. The most active project have a new release every 6 weeks or so. (This certainly struck a chord with me.)
This brought us to the second part of Barry's talk on Software as a Service (SaaS). The point of a SaaS system is to offload the user from the responsibility of building and maintaining an IT system. The job of building a SaaS system is integrating a lot software packages to provide the service. SaaS also has to deal with transparently upgrading the service to the users as it implements new features and fix bugs. As such it complements Open Source and it rapid development cycle.
SaaS is a newer market and there are only a few emerging SaaS BI services available now. Barry touched on three, LucidEra, SeaTab and Oco, all early stage start ups. There are some real architectural challenges to providing BI as a service. For example, one issue is security. For a number of reasons, many Open Source projects put security on the back burner. On the other hand, a Saas customer needs to have solid assurances that their data is secure and safe from other customers of the SaaS service.
We will have to see how these trends play out. Barry quoted a research report that suggested that software innovation goes in a roughly 15 year cycle, and that in Business Intelligence we are just entering a new cycle that can be expected to go on to 2020 or so.