There’s something surprising about enterprise technology: It’s not as evolved as most people assume. According to Flexera’s 2021 State of the Cloud Report, just six of 10 businesses have migrated their workloads to the cloud. That implies that 40% of workloads are still running on a client server or mainframe infrastructure. But not for long.
Jamie Dimon, CEO of JP Morgan Chase, recently announced that his company will invest $12 billion in technology this year, a 26% increase since 2020. The aggressive investment will continue the bank’s digital transformation, increasing and improving digital products and services to compete with a slew of new fintech startups.
As a result of Chase’s growing portfolio of digital products and services, the company will have to radically transform its operational methodologies to function less like a bank and more like a software company. This means: Chase must shift from monitoring its infrastructure to monitoring its applications. After all, the performance of the company’s apps will almost entirely dictate the user experience.
Many enterprises will likely follow a similar, albeit less expensive, path as Chase. And they too will have to focus on app availability as they offer more digital products and services on more ephemeral machines.
There’s only one way to improve service assurance in such modern IT environments: AIOps (artificial intelligence for IT Operations). AIOps tools ensure availability by helping DevOps and site reliability engineering (SRE) teams quickly identify and fix issues that affect the performance of an enterprise’s apps and vital services.
But it’s not just about buying an AIOps platform – it’s about implementing the right AIOps platform. In our increasingly app-focused world, AIOps tools that speed application fixes will be more helpful than those supporting infrastructure and traditional IT operations.
It’s All About Time
Effective AIOps tools mitigate the cost of application downtime. And these costs can be staggering. Most large enterprises say just an hour of system downtime can cost over $100,000.
There are other costs too. Most of the working world vividly remembers when Slack went down at the beginning of the year. The hours-long outage, sparked by AWS network scaling issues, undoubtedly cost the company money. But there were additional repercussions. Slack stock dropped and bored Slack users flooded social media with commentary about the outage.
In the old days of events monitoring, engineering teams wouldn’t get event notifications of a problem until something went wrong. They’d spend significant amounts of time chasing after an incident, trying to figure out what caused it and racing their way toward a fix. But modern enterprises simply can’t afford the costs associated with operational disruption, dissatisfied customers, diminished sales, and plummeting stocks.
Now AIOps tools gather metrics, traces, and logs so early in the incident lifecycle that they can capture incidents before they occur. And, unlike older, rules-based technologies, these algorithms work well on partial evidence, which speeds detection of potentially disruptive incidents. The time saved is invaluable, as the sooner incidents are spotted, the sooner DevOps and SRE teams can get to the root cause, prevent an outage, and sidestep damaging downtime.
A New Definition for ROI
Given the cost of downtime, an AIOps tool’s return on investment is almost certain. But modern companies should look beyond a return on investment to a return on innovation.
If AIOps guarantees application availability, the tool also protects the customer experience. This creates an environment where SRE teams and DevOps practitioners feel safe to roll out innovations. Knowing they can seamlessly implement updates without interrupting service, teams can deploy new applications into production environments at increasing velocity. And these innovations are a critical element of success in our digital economy.
In addition to helping teams break free from risk aversion, AIOps tools automate the toil out of the working days of SRE teams and DevOps practitioners. In the pre-AIOps world, teams would have to string together legacy monitoring tools and manually go from one tool to another to pursue an incident and follow it through to resolution.
AIOps frees engineering teams from this labor-intensive process. True AIOps solutions smoothly interlink data from across an organization’s IT operations and automate processes to ensure top performance of apps and services with minimal human interaction. The tool continuously collects data, scanning it for performance affecting issues. As a direct result of this automation, SREs and DevOps practitioners can focus on the continuous development that generates revenue and brings enterprises real value.
Companies are taking notice of both the cost savings and the accelerated innovation advanced AIOps tools enable. A few weeks ago, I met with a senior vice president of AIOps for a top-five global bank. She is championing AIOps technology internally as the bank becomes application-oriented to compete in the fintech market.
The very existence of a high-level AIOps role speaks volumes about where the technology is going. When I first entered the AIOps market in 2012, I couldn’t get anyone to talk to me about this exciting new AIOps technology. Indeed, back then, it wasn’t even called “AIOps”! But as enterprise technology evolves and companies respond to consumers’ growing appetites for bigger and better apps, AIOps is becoming essential to availability and innovation. And it will soon be a cornerstone of every company’s digital strategy.