Information technology has morphed into a beast that companies 10 years ago couldn’t have imagined. Multicloud, distributed systems, software-defined everything — it adds up to a circus of unruly animals. The old methods — human surveillance and traditional monitoring — can’t tame them all. Companies need both human and artificial intelligence to navigate the new IT maze. And the latter may not exist in adequate measure in some obvious product choices on the market.

Human brains stand no chance of monitoring all the events occurring in an IT environment today, according to Phil Tee (pictured), co-founder and chief executive officer of Moogsoft Inc. Before cloud and virtualization, a high event rate would have been 100 to 200 events per second, he added. “Nowadays, with an average customer of ours, you add a zero or two to that rate — maybe even three,” Tee stated.

Legacy systems struggle to ingest and interpret all of that data. Newer technologies with machine learning and artificial intelligence can sort through it and preempt problems. AI for IT operations, or AIOps, looks for factors in the data that could impact service or cause an outage. It leverages a number of AI types. Deep learning and convolutional neural nets are trained to recognize features of data that represent some known factor. Feature detection, which looks for patterns that are not obvious to the human eye, can catch problematic events before they disturb business.

Moogsoft’s AIOps platform cuts through data noise to locate and fix issues quickly, according to Tee. The company claims it’s invested in leading-edge technologies that some well-known competitors — like Splunk Inc. — can’t compete with.

Tee spoke with John Walls (@JohnWalls21), host of theCUBE, SiliconANGLE Media’s mobile livestreaming studio, and guest host Justin Warren (@jpwarren), chief analyst at PivotNine Pty. Ltd., during AWS re:Invent in Las Vegas. They discussed the need for AIOps in modern IT and Moogsoft’s quest to mainstream the freshest AI tech. (* Disclosure below.)

This week, theCUBE spotlights Moogsoft in its Startup of the Week feature.

It’s all better with AI

Companies are slathering AI like ketchup on just about every technology product that exists lately. Data cataloging, management, monitoring, customer-contact interfaces — they’re all better with AI. Without a doubt, some of this is hype mongering; real end-user benefits may not be much to jump up and down about. Does AIOps really deliver more than all those other IT monitoring tools on the market?

“There are some purists who claim there is no such thing as AI, and everything that we talk about today is really machine learning or data-driven algorithms,” Tee said. There is some truth to that. However, there is real value in the AI and ML available today, he added. It can and does allow humans to do things they couldn’t do by themselves.

Revenue generated from direct and indirect application of AI software is estimated to grow from $643.7 million in 2016 to $36.8 billion by 2025, according to research from Tractica LLC.

AI in IT operations is producing new, intelligent systems that lift the burden from admins, according to Chris Scott, managing director, Accenture AWS Business Group, at Accenture LLP. “These systems listen, they see, they can adapt, they understand what’s going on just like people do,” he recently told theCUBE.

Seeing farther than supervised ML training

Some of the most impressive work that AI can do is detect the “unknown unknowns,” Tee pointed out. This is especially helpful in monitoring vast, distributed IT systems constantly choking on a deluge of new events.

In the past, monitoring systems had lists of things that might be ignored or not depending on rated severity, etc. This differs from the information science that Moogsoft applies to monitoring.

“We can measure the semantic content and the informational content of events to work out whether it’s telling us something of import,” Tee said.

This technique can eliminate 90 to 95 percent of inbound data as effecting nothing. That narrows the data inundation down to the point where Moogsoft can process it in real time with compute-intensive AI algorithms. They can accurately indicate instances or potential instances, according to Tee. 

AI and ML techniques that alert based on supervised training and what’s already known only go so far in instance detection. “That’s why we use a combination of unsupervised feature detection and supervised learning,” Tee said. “The unsupervised feature detection just knows something as an unusual, highly correlated pattern or feature in the data and needs no prior understanding of what’s going on.”

Moogsoft spanks competition with unique IP

Moogsoft wants to shorten the path from the invention of new AI techniques to enterprise adoption. Unlike other software companies that may be repackaging what is available in open source, Moogsoft is investing in unique IP. The company’s enlisted Ph.D. scientists, and holds 44 patents currently, according to Mike Silvey, executive vice president of business development and product strategy at Moogsoft.

“Our perceived competitors out there, folks like Splunk and ServiceNow — no investment in IP,” Silvey told theCUBE. “They’re trying to take old technologies and old techniques to solve a problem that they just can’t solve.”

Are there any proof points that Moogsoft delivers better protection against costly outages? Silvey offered an anecdote.

Last year, an American airline suffered a six-hour outage. Its ground-handling software failed, and it could not schedule a single flight for that period. “This year they have Moogsoft. And our software detected an incident that they could action earlier, resolve before it impacted their ground-handling system,” he said.

Detecting that unknown prevented an outage of flights across the U.S. that would have lasted a minimum of 4.5 hours, according to Silvey. 

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of AWS re:Invent. (* Disclosure: Moogsoft Inc. sponsored this segment of theCUBE. Neither Moogsoft nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE