Data Liberal or Conservative

Guest Blog – IoT: Are you a “data liberal” or a “data conservative”? June 23, 2014 | by

by Andy Thurai, Guest Blogger

In the last decade, as a society, we had worked very hard toward “liberating our data” — unshackling it from the plethora of constraints unnecessarily imposed by I.T. In contrast to this, In the 90s and early 00s, data had been kept in the Stygian depths of the data warehouse, where only an elite few had access to, or had knowledge about it or the characteristics defining it.

Once we had the epiphany that we could glean amazing insights from data, even with our “junk” data, our efforts quickly refocused around working hard to expose data in every possible way. We exposed data at the bare bones level using the data APIs, or at a value added data platforms level, or even as industry based solutions platforms.

Thus far, we have spent a lot of time analyzing, finding patterns, or in other words, innovating, with a set of data that had been already collected. I see, however, many companies taking things to the next proverbial level.

In order to innovate, we must evolve to collect what matters to us the most as opposed to resign to just using what has been given to us. In other words, in order to invent, you need to start with an innovative data collection model. What this means is for us to move with speed and collect the specific data that will add value not only for us, but for our customers in a meaningful way.

This is why we are seeing a proliferation of devices/sensors everywhere. Name any industry segment and you’ll discover that within that segment waves are being made with the word “smart” added to the intuitive data collection initiatives.

There are smart-cities, smart-grids, smart-homes — even smart toilets. Essentially, we need to connect the data collection devices to the Internet to easily collect the data. To some extent we bypass the privacy, security and sensitivity of these data collection points. These IoTs (Internet of Things) are in that sweet spot of “tug of war” between “data conservatives”, who are afraid of opening up their kimonos and who try to stall things using security and other scare tactics — and who want you to guard your data in Ft. Knox, and the “data liberals”, who want to open everything up and innovate everything to make life easier for everyone.

It’s true that all of these data collection initiatives will get you closer to the source of “all things data” with respect to helping you collect — what you want, when you want it. However, in the process, you may end up magnifying secondary issues you aren’t even cognizant of such related to: data transportation, data qualification, data integration, and, my favorite, data security.

True – security, privacy and compliance will be hot topics for the foreseeable future. But, as with any maturing technology, if you wait for the most secure, most mature, and most innovative platform to arrive, you will be left in the dust with an extinct data dinosaur sooner that serves none of your interests.

There is an alternative. Companies offer platforms, called IOT as a service platform and IOT middleware platforms. Most recently, I had the opportunity to take a closer look at one such platform – Xively (Owned by LogMeIn.) As Xively calls it, “harness[ing] the power of IoTs quickly and easily to transform your IoT vision into a market reality.”

What I really like about them is that Xively offers solutions that don’t require you to take sides in this highly polarized debate. They offer (web)sockets, REST and MQTT choices (I am sure they will add when a new version shows up) for connectivity choices, as well as a completely secure platform using a combination of SSL/TLS and API keys and a private cloud, and integration with most data formats such as JSON, XML, as well as non-structured data formats.

What impressed me the most is their “Flexible Data Service” which is a high performance time-series database. For those who are neck deep in IoT enabled real time applications, this solves a major issue of tracking your data, time enabled. Not only is that provided out-of-the-box, but you can also set up a series of actions and triggers based on a time series of data across any connected device/service/application. What a powerful feature! Xively essentially takes the dependency out devices, sensors, connectivity & manufacturer elements, and allows you to build a truly neutral IoT platform to suit your needs.

In this reality vs dreamy eyed fantasy initiative, I submit that the security background in me indubitably sides with the data conservatives. And yet the inventor in me sides with the data liberals. Now it’s your time to make a decision. Who will you vote for?

You can always vote neutrally by staying with a platform such as Xively. Check out a detailed offering here You won’t be disappointed.
Andy Thurai Andy Thurai is a technologist turned story teller. He is an emerging technology strategist, solution specialist, thought leader, speaker, recognized author, blogger, evangelist and an expert in selling software to enterprises.  His interests and expertise include IoT, API, Big Data, Analytics, Cloud, Mobile, SOA, identity management, security, governance, and SaaS.  He has held multiple technology and business development leadership roles with large enterprise companies including Intel, IBM, BMC, CSC, Netegrity and Nortel for the past 20+ years.

He blogs regularly at on various topics. You can find him on LinkedIn at or on Twitter at @AndyThurai. He also provides IoT consulting, advisory and security services at



One Response to “Guest Blog – IoT: Are you a “data liberal” or a “data conservative”?”

  1. alex says:

    Good points that apply across many businesses. I am into Personalized Medicine, where Big Data collections are becoming possible and necessary to help solve the complexity of human populations. My only comment to your list of concerns is cost. Security, privacy, compliance are important as mentioned. GWAS DNA sequencing costs are dropping considerably, the human genome is becoming cheaper to obtain. However, collecting and comparing data, along with other phenotypical and molecular data is costly and not without controversy. Equally important is data ownership, data sharing, quality control. I vote liberal, with a hint of conservativism.