AIOps adoption has grown rapidly since Gartner Research defined category. Let's take a look at the customers and vendors embracing the movement in 2018.
Not much longer than a year ago, Gartner Research introduced the AIOps category (Artificial Intelligence for IT Operations), and a lot has changed since then. In a nutshell, the AIOps category captures the emergence of products that use artificial intelligence and machine learning to assist humans working across IT Operations functions, including availability and performance monitoring, event correlation and analysis, IT service management, and automation.
Moogsoft was one of the first vendors to self-identify within the AIOps category, and naturally, we took a lot of flak for that. Initial feedback that I received from enterprise buyers, partners, and competitors were along the lines of, “this space is too immature,” “AI & ML are years away from actually working,” and “the ITOA category isn’t going anywhere.”
Considering it’s roughly a year down the road, I decided to take a look at AIOps adoption across the market. I looked into which enterprise organizations are embracing AIOps today, who are the customers, and who are the vendors.
Here are my findings.
AIOps is Spreading Beyond Traditional ITOps
After a little research, it’s clear that AIOps is beginning to go mainstream. AIOps functions are now spreading beyond traditional IT Operations roles, and into specialized data science roles.
Companies are now building their next generation of applications and services with customized instrumentation to generate rich data while maintaining strong signal-to-noise ratios. This live streaming data across various sources of a microservices ecosystem can then be centralized, enriched with available configuration data, and then analyzed and correlated in real time for predictive and proactive indications of failure. Automated alerting can intelligently pull in relevant resources to address the issues, and ideally make iterative improvements in resiliency and self-healing.
Enterprise IT Organizations are now introducing Data Scientists with AIOps specializations to make these visions a reality.
See a few examples of active job descriptions below:
Position: Sr. Data Scientist
Description: The AIOps Data Scientist will be focused on solutions to correlate, consolidate, alert, analyze, and provide awareness of events within the global enterprise infrastructure, network, data centers, and cloud environments. Identifies ways to leverage data to answer important business questions. Innovates applications and techniques to use and communicate data in new and more effective ways for influencing business decisions, encouraging long-term customer engagement, enhancing security, and any other possible uses.
Position: Data Scientist
- Contribute to delivering transformative AIOps solutions on our SaaS platform
- Work with customers and product experts to blaze a trail in autonomous data center management
- Solve customers’ complex IT problems using innovative analytical processes and techniques
The AIOps Vendors for Event Correlation & Analysis
While homegrown AIOps projects are increasing across Enterprise IT, a more popular option is to leverage third party, purpose-built AIOps platforms.
- Building integrations across toolsets is extremely difficult
- Building highly scalable platforms is extremely difficult
- Building real & automated machine learning algorithms that actually work is extremely difficult
Let’s assume that the three points above are achievable. Still, the associated cost and time required to get a homegrown AIOps platform into production likely aren’t worth it. A Fortune 500 Financial Services organization financially justified purchasing and deploying Moogsoft AIOps across their entire IT infrastructure by identifying that the time to get Moogsoft AIOps into production was two years earlier than the homegrown project they had been pursuing, while the cost was just a fraction.
Gartner’s Market Guide for AIOps Platforms, published in August of 2017, includes 20 AIOps Vendors that are rated with varying degrees completeness. The list includes companies like Moogsoft, BMC, IBM, Splunk, Rocana (assets now owned by Splunk), Elastic, SumoLogic, Loom Systems, and more.
From this list of 20 vendors, there are roughly four vendors that I see getting adopted by enterprise organizations for AIOps event correlation and analysis projects today. I decided to do some research and provide a summary around areas like company background, strategy, and product.
|Vendor||Year Founded||Employee Count||AIOps Product||AIOps Product Launch||Touted AIOps Customers|
|BMC||1980||~6,900||BMC TrueSight Operations Management||2015||Park Place Technologies, Sanofi, inContact|
|IBM||1911||~380,000||Netcool Operations Insight||2015||Austrian Federal Railways, Claranet, The Plastic Bank|
|Moogsoft||2012||140||Moogsoft AIOPs||2012||GoDaddy, SAP SuccessFactors, Royal Bank of Canada|
|Splunk||2003||1,000-5,000||Splunk IT Service Intelligence (ITSI)||2015||Vodafone, Cox Automotive, Molina Healthcare|
|Vendor||Target Market||Touted Use Cases||End Users|
IT Ops across Traditional IT
– IT Service Management
– Cloud Management
– Workload Automation
– IT Automation
IT Ops across Traditional IT
IT Ops and DevOps across Traditional, Digital, and Hybrid IT
– IT Automation
– Incident Management
IT Operations, Security, and Business Analytics across Traditional IT, Digital, and Hybrid IT
– IoT & Industrial
– Business Analytics
– App Delivery
– Solution Engineers
– Data Scientists
|Vendor||On Prem or SaaS||Minimum # of Products Required for AIOps||Product Release Cycles|
(based on 2017 releases)
|# of Supported 3rd Party Integrations||Trial Offered|
|BMC||On-Prem||3||10 Months||6||Yes – 30 Days|
|IBM||On-Prem||13||No update since V184.108.40.206 (June 2016 GA release)||123||No|
|Moogsoft||On-Prem & SaaS||1||Bi-Weekly||45||Yes-30 Days|
|Splunk||On-Prem & SaaS||2||3 Months||262||Yes-7 Days|
|Vendor||Analytics: Real Time / Storical||Noise Reduction||Event Correlation||Service Mapping||Probable Root Cause||AIOps Mobile|
-Dynamic Thresholding (time-series)
– Fuzzy Matching via Language
-Fuzzy Matching via Time/Seasonality
– Fuzzy Matching via Time, Language, Topology, Neural Feedback
– Service Mapping
– Rigid Matching via Time & text
AIOps is Set for Growth
Gartner claims that, “By 2019, 25% of global enterprises will have strategically implemented an AIOps platform supporting two or more major IT operations functions.” Based on my observations on AIOps adoption, this number sounds realistic.
With growing AIOps budgets, specialized jobs, and proliferation of vendors going after the space, there’s no doubt that 2018 will be a pivotal year for AIOps across Enterprise IT. Whether it’s by building homegrown projects to intelligently create, analyze, and take action on operational data, or by purchasing and growing the number of purpose-built solutions, AIOps should be a part of every IT professional’s vocabulary by the end of the year.
About the author
Sahil Khanna is a Sr. Product Marketing Manager at Moogsoft, where he focuses on the emergence of Algorithmic IT Operations. In his free time, Sahil enjoys banging on drums and participating in high-stakes bets.