AIops speeds IT discovery and troubleshooting, but it isn't magic. And operations staff must still prepare the data for machine learning and manually solve some issues.
Artificial intelligence and machine learning can slash the number of false alerts that tie down operations staff, speed troubleshooting of problems, and help developers and architects understand and manage fast-changing, cloud-based IT environments.
But CIOs should not expect what some customers call “magic” results, such as automatically predicting and fixing any conceivable IT issue, or even just accepting any log or event steam and analyzing it without any data cleansing or normalization.
AIops is the use of artificial intelligence to manage, optimize, and secure IT systems more quickly, efficiently, and effectively than with manual processes. Market researcher Gartner estimates that the AIops market ranged between $900 million and $1.5 billion in 2020 with a compound annual growth rate of around 15% between 2020 and 2025. Along with standalone AIops platforms, many IT observability, management, and monitoring tools integrate with AIops platforms or have added AI capabilities to their products.
AIops is best, according to customers and analysts, at quickly scanning massive amounts of data from hundreds or thousands of sources to filter out the most important alerts or identify underlying trends, as well as quickly detecting new elements such as application programming interfaces (APIs) that link applications— those “things that human intelligence can no longer handle,” says Sean Mack, CIO and CISO at Wiley, a global leader in research and education. It is ideal, he says, for providing insights into IT issues among “the exponential growth of the complexity of our systems and services,” with virtualized elements that “may be there one second and may not be there another second."
But AIops efforts can fail if businesses don’t understand its limits.
Continue reading with a free subscription to CIO.com.