Blog

Let Your Date Take You To The Next Level

Data

Is your business taking advantage of the data revolution that is going on right now? There are so many really exciting technologies that have become mainstream in the last couple of years that are truly revolutionizing data management and you should be taking advantage of them. They can make a real impact on your business. Let's look at some of them:

Database As a Service

What a concept! Who needs the database? Just get the database functionality!

Just a short time ago, if your business required you to collect data, you had to purchase a database software like Microsoft SQL Server, Oracle, etc. and a server to run it on which meant a piece of hardware and an operating system. In addition to the operating costs like electricity and possibly cooling, you also had to worry about maintenance of the server hardware, the operating system and the database software, all of which required regular maintenance in the form of firmware updates, software updates, and security fixes. I'm not even mentioning all the upfront costs you had to incur purchasing the hardware, the operating system and database software licenses. After all this, there were yet more costs involved. For example, you would run out of hard drive space on the server and end up upgrading the hard disk or at some point you needed more memory to improve server performance. All this trouble just to have database functionality!

Now, you can just get the database functionality from a cloud provider. For example, Microsoft offers SQL Server functionality through Azure SQL Database offering. No hardware to purchase, software to install or maintain. Let experts -- who are truly experts -- take care of all the maintenance for you. There are no upfront costs and because this is a metered service, you only pay for what you use.

NoSQL Databases

For many years, when someone said database, it meant one thing: a relational database. The only thing you needed to decide was which relational database to use e.g. Microsoft SQL Server, Oracle, MySQL, etc.

Relational databases work great for what they're designed to do but they have their short comings. They don't scale very well and strategies to scale relational databases are not for the faint of heart. Doing analytics using relational databases could also be a bit challenging and may require manipulating and sometimes denormalizing your data. Hearing the words relational database and denormalized data in the same sentence sounds down right blasphemous.

In recent years though, NoSQL databases like document databases, key-value stores and graph databases started to gain momentum and they truly provide functionality that should be a part of your data management strategy. As a matter of fact, I would even argue that if a NoSQL database is not part of your data management today, you probably should re-examine your approach to managing your data.

Insights, Analytics and Machine Learning

For a lot of businesses, data analytics means generating some a la carte reports. Also, getting into meaningful data analysis requires both expertise and computing power that may not be available at every business.

This is where a business should utilize both the cloud and outside expertise to dive into its data. What's interesting about data is that it is multi-dimensional. When analyzed the right way, your data may produce 2 + 2 = 7 because you're discovering and linking seemingly loose and unrelated data which may end up producing truly unexpected results.

For example, I recently watched a video on YouTube of an expert demonstrating machine learning features that are available on Microsoft's Azure cloud platform. In the video, he had over 32 thousand records of data which included information about people's age, location, gender and income level to name a few data points. He then tells the system to use 80% of the data to "learn" the correlation among a person's age, location and income. After a little bit of data crunching, the system gets done "learning". The expert then tells the system to use the remaining 20% of the data to check the accuracy of its learning. Keep in mind, the machine learning system never looked at the 20% of the data that it will now use to check the accuracy of its learning. The system goes to work again and in a short while, it comes back to indicate that it has a pretty high percentage of accuracy in correctly "guessing" a person's income level based on his/her age and location.

This is amazing technology that is now available to any business. Doing something like this just a few years ago for a small to medium size business would have been unimaginable.

I suggest, you jump on the bandwagon now. There are some really exciting things happening in data management technologies that can have a meaningful impact on your business.