What IBM’s Bluemix means to big data and analytics
Throughout the 2014 IBM Impact conference I am drawn to the prophetic rise of the software developer in Stephen O’Grady’s book, "The New Kingmakers.” While the entire book walks though the new world order of the empowered developer, a comment in the introduction of the book, and a deft diagram all channeling Billy Marshall’s comments in 2008’s “The CIO is the last to know” stand out. With the advent of cloud availability and the increasing number of technology services in the cloud, developers (and soon to be developers) are deciding on the modern organization’s technology platform, not the IT department. This significantly affects the way we all (including IBM) do business. Enter Bluemix.
Bluemix enables developers (read, DEVELOPERS) to rapidly build, deploy and manage their cloud applications, while leveraging a growing ecosystem of services and runtime frameworks, all on demand and all elastic. Some of the existing services that we have had in the beta have been: MapReduce (the same goodness from Hadoop), BLU Aceleration, SQL Database, NoSQL (JSON Cloudant CouchDB) Database as well as a dedicated database service for Mobile applications, offering just enough database without the clutter.
So fast-forward to this Tuesday’s (4/29/14) announcements at IBM Impact 2014. We are adding a slew of additional big data and analytical services to Bluemix including:
- Geospatial Analytics: Built on InfoSphere Streams, this service enables real-time region monitoring and produces notifications and alerts when a device enters (or leaves) a pre-defined region.
- Time Series: Allows developers to manage a data store for “Internet of Things” device data, perform analysis on that data and minimize storage used. We are talking a data service made for sequential data such as that generated from smart meters, RFID tags and other sensors and devices.
In addition, Bluemix is also adding new services for Reporting (think Cognos with enhanced visualizations), Internet of Things and even a service capability of Watson Analytics, which accepts natural language queries to return applicable datasets and visualizations in common (non-statistician) lexicon.
Ok, great—more stuff. So what does this mean to the developer?
Just about everything.
These announcements signify a massive accumulation of big data and analytics technology in the cloud, as a service. By having these solutions built as services to the IBM Cloud, developers no longer have to deal will installation issues, hardware and software configuration issues or other similar issues—it is all available the in the cloud for you.
The best way that I visualize the utility of Bluemix is as cooking dinner for the family. In the past, if you had a number of dishes and recipes you would have to go to the market (or markets) and buy the groceries, spending a decent amount of time getting to the store, finding parking, burning gas (for those of us outside of the city) finding what you need and then bringing it all back home. Then, once home, you have to clean and cut the vegetables, prepare the sauce, boil the water, marinate the meat and so on—you get the picture.
With Bluemix, everything that you need is right there in front of you, as you need it, already washed, chopped, marinated and ready to go. What if you change your mind on the main course or forgot an ingredient? No worries—it is all right there in front of you. No trip back to the market, no delay.
Bluemix is IBM’s developer construction kit: pretty much anything that you need to build an application, right at your fingertips.
For big data and analytics in general, this all means that more people will be learning, experimenting with and ultimately using the latest in technology to glean insight from data. The fact that all of this is in Bluemix as a service (rather than in a traditional individual installation) removes much of the barrier to entry, especially with some of the latest Hadoop and Streams technologies. In addition, having it as a service in a cloud environment empowers developers to add, remove and test different capabilities based on the project’s needs, even opening the door to incorporating multiple database types, each one used based on the workload type.
Look, I’m not saying that Bluemix will remove the demand for an on-premise solution, but it will, at a minimum, drastically reduce the time and cost for developers to get an application off the ground. It will also open new doors for trial and the ability for developers to expand their portfolio of skills.
So let’s get back to my original thought around the CIO being the last to know. As the market shifts to a developer-led economy, organizations are going to have to deal with an influx of new technologies and skillsets, both self serve and good enough, but they should not be so nervous. While the technology itself needs to be vetted for security gaps and governance issues, since the developers are ones searching out and building upon this new technology, they are also bringing the skill set as well. Think of the old days where the company declared a platform or language standard: they would have to train or find the developers to fit their requirements. With the developer led economy and a developer construction kit, your development team is already self-educated and empowered.
With these recent big data and analytics services being released into Bluemix, it is only time before we see a new revolution in analytical applications.
Go ahead, give it a try and let me know what you think.