How Artificial Intelligence is Humanizing Big Data

Artificial intelligence (AI) has progressed by leaps and bounds in recent years. For many people, it is a subject that still holds a science fiction-esque mystery and a good deal of skepticism. But on a day-to-day basis, almost everyone who uses a computer or turns on a light switch is directly interfacing with or using a system supported by a form of AI.

Logz.io, a company I covered a few months ago, announced a major development in the AI space this week that could totally change the information technology (IT) space.

Anyone who uses Gmail is likely familiar with a form of AI that Google uses to track events and send its users helpful alerts. If a Gmail user gets an email with a flight confirmation, Google’s AI will read that email, extract the important information, and add that event to the user’s calendar. The first time this happens can be a little off-putting, raising questions like: How did Google know what I was doing? Did someone read my email without my permission?

In reality, the technology behind that feature is completely innocuous (and can be disabled). However, it is also extremely powerful. The concept, which Logz.io is replicating in the IT space, is that a machine can be taught to read, think, and process very complex information.

The AI development that Logz.io announced this week is called Unified Machine Intelligence (UMI™), which powers a broader platform called Cognitive Insights™. This program is trained to teach itself incredible volumes of information about IT management.

For example, UMI can scan the internet for forums where people discuss problems they have had with certain programs, errors they have received, and solutions they have discovered. UMI can read, capture, and process that information, in essence teaching itself IT management. The implications of this are substantial and numerous.

To begin with, Logz.io believes UMI can become the most educated IT manager (human or machine) in the world. Tomer Levy, CEO and Co-Founder, had this to say about its intelligence potential, “Exponential learning is not something that humans are capable of doing. We have the advantage of being able to learn, and learn how to learn things we do not know yet, but the ability to study, store, and automatically access such immense volumes of data is an advantage of AI.”

With a few basic skills, UMI is able to endlessly grow its knowledge base, and it gets smarter by the second. But UMI also learns through day-to-day activity.

As companies use Cognitive Insights, they interface with huge volumes of data about their IT environments. They will look for bugs in their system, find them, fix them, and move on to perform other activities. All the while, UMI is learning that human behavior. It is learning what humans see in the immense universe of log lines (single lines of code generated by any action on a computer or network), and sharing that intelligence with other companies who are using the platform.

For example, if an application crashes and the IT manager logs into their system to find the problem, they will be faced with tens of thousands of lines of code in which they have to find the one that is relevant. It is almost the definition of finding a needle in a haystack. Or as Levy puts it, “finding a needle in a needle stack.”

But UMI has seen that event take place before for another company, and the IT manager at that company marked it as being an important issue. So UMI will highlight that event automatically and supply helpful resources that it has found from its research online for solving the problem.

Every event that goes through the system, and every time a human takes an action within the IT environment they are managing, the smarter and more helpful UMI should become.

Today, companies are inundated with mountains of data. While helpful to have, it can be a hindrance without the ability to highlight the data that matters, and only very small pieces of it matter. Without AI systems like this one, there is no structure that can be used to humanize data. But by leveraging AI, by teaching it how to be human in a sense, suddenly data can be quickly sorted and managed.