By Antonio Piraino

LONDON, UK - The idea that artificial intelligence can effectively assist with cognitive tasks is well-established in the world of business. Automation in various forms is already seeing significant use in back offices, with all manner of routine administrative errands – data entry, invoice filing, CRM updates, claims processing, and so on – up for grabs. Finance, accounting, and customer services departments appear particularly primed for disruption.

The use of artificial intelligence in IT departments isn’t necessarily what first springs to mind, and yet IT is precisely where automation already has a significant foothold. Gartner has coined a term to describe the phenomenon: “AIOps”.

Solving the problem of ever-increasing workloads with AIOps

AIOps is a solution to a long-standing problem in IT operations, namely, the problem of ever-increasing workloads.

With the rapid onset of digitization, business' technology needs have grown exponentially over the last decade or so. Staff are expected to manage more than ever before - more services, more applications, more incidents. Quite understandably, they're struggling to cope.

Aggravated by the growth of the Internet of Things (IoT) over the last few years, staff are now overseeing significantly more data than before. According to IDC, there is expected to be 163 zettabytes of data produced annually by 2025 as the IoT takes off. That's ten times the amount we currently produce (16.3 zb). ITOps staff simply can't keep up with these volumes and can't analyse the inbound information streams in a cost-effective or time-efficient fashion.

Business IT is now more complex than ever. There's too much to do, and too little time with which to do it all.

Related: Why MLOps Is Your Business' New Competitive Frontier

AIOps for IT operations

Cue the birth of AIOps; IT operations platforms that, unsurprisingly, use artificial intelligence. Powered by machine learning, AIOps analyses big data from various IT operation tools and devices to spot and - even more impressively - react to issues in real time.

What does this mean in practice?

Let's imagine an application failure. An IT ops staff member walks over to their computer screen to see thousands of messages logged within a specific timeframe - a nightmarish situation, but a depressingly common one.

It's anyone's guess what exactly has gone wrong. The critical event that triggered the log entries is embedded in the data somewhere, but a dataset like this is impossible for a human being to review line by line. What are they to do in this situation?

AIOps is designed precisely for these situations. AI is better equipped to both gather data and to filter out any irrelevant pieces of data, particularly when that is spread across a range of devices and applications.

With only humans in the equation, it may take hours to locate and solve an application failure. AI, on the other hand, may take just a few seconds. Liberated from firefighting IT problems, staff can undertake more proactive, meaningful work.

Looking to the future of IT problems

AIOps is a rapidly developing area. As things currently stand, AIOps allows for partial automation: some tasks are automatic, reducing the average time it takes to resolve issues, but humans ultimately still form an integral part of monitoring for issues, and fixing problems when they arise. Current AIOps systems struggle to understand the relationships between applications, infrastructure, and other datasets.

Expect that to change in the next two or three years. As automation technologies evolve, monitoring will become mostly - or even fully - automated. IT problems will also increasingly be solved automatically, resolved before businesses even notice there was a problem in the first place.

Looking to the future, as AIOps technologies develop and become more impactful in creasing efficiencies, businesses will find it harder and harder to sustain their competitive edge with performance bottlenecked by outdated and manual work. By applying AI to IT operations, IT issues become easier to identify, predict, prevent, and even fix.

As workloads for IT staff swell, organisations will be pressed to select an approach. They can capitulate to information-overload, or begin taking measures to harness the tremendous opportunities that AI can offer.

Antonio Piraino, Chief Technology Officer at ScienceLogic.