By the time a human manages to warn you about a risk to your company’s cloud program, it’s often far too late to do much more than tally up damages and regroup. Whether you’ve just been told that your cloud infrastructure is starting to fail, that your company’s cloud migration project is months behind schedule, or been handed an eye-watering seven-figure cloud services bill, it’s clear that “human speed” is no match for cloud’s capabilities. To properly manage company cloud initiatives with increased speed, adaptability, and detail, leaders now leverage artificial intelligence (AI) and machine learning (ML) technologies, hoping to fill in the gaps that legacy (human) processes don’t or can’t address.
Any heavily automated process tends to shine when properly complemented with AI integrations. AI allows you to smartly automate certain very intensive tasks, freeing up more of the workforce to move from focusing on grunt work to functioning in a more strategic capacity. When combined with the cloud’s considerable improvements to workload efficiency and scalability, AI-derived automation allows IT teams to spend more of their time generating organizational and business value, and less time babysitting basic-but-intensive processes.
AI to Strengthen Network Infrastructure
Recently Accenture partnered with a leading Asian energy services company, with the goal of helping to improve the company’s existing fraud response systems. From the outset, it was quite clear that sole reliance on human-centric processes was no longer sufficient to properly monitor the network comprising over 600 hosts, 3000 networked devices, and 1500 Internet of Things (IoT) enabled devices deployed across some 700 physical retail locations. Accenture rose to the challenge, of designing and implementing a new system based on its own AI-enabled process automation platform, myWizard, to radically transform the company’s cloud and networking landscape. Equipped with brand-new self-healing functionality, and by letting AI do the heavy lifting of both monitoring infrastructure and handling scheduling/notifications, the collaboration was a resounding success. Since the collaboration began, fraud monitoring efforts now run 24/7/365, and fraud service response coverage alone has improved by ~80%.
[RELATED: Artificial intelligence needs human judgment to work.]
AI to Master Cloud Migrations
Another large slice of “life in the cloud,” as it were, involves cloud migrations. Often lengthy and technically challenging to navigate, even the most well-planned migration can morph into a costly, off-track, migraine-inducing mess – and for some companies, that’s enough to stall their cloud journey altogether.
No matter the company size or industry, cloud migration challenges tend to crop up. CloudEQ, a cloud consultancy firm concerned with empowering digital transformation through well-managed cloud migration, discovered this when partnering with multi-cloud management AIOps platform vendor Virtana to tackle a global fast-food retailer’s impending, massive cloud migration effort. Workloads in North America had been migrated previously and were already operational, but those processes had taken too much time to repeat — and with the rest of the globe still very much on the “to-do” list, the chain knew it was time for experts to step in.
It was a good call. As Sean Barker, CEO of cloudEQ identified, “Many of this company’s restaurant locations are open 24 hours a day, seven days a week. With that kind of operating environment, they can’t afford to be down a single minute.” Realizing that only a custom solution could deliver that uptime was a key part of the strategy; the remainder, as derived from AIOps-enabled insights provided by Virtana, involved carefully untangling complex application dependencies so as not to disturb any on-premises processes. Happily, both partners were more than equal to the challenge, with the resultant cloud migration not only delivered ahead of schedule but alongside around $750,000 in recovered cloud costs.
AI for Cloud Cost Management
Cloud cost management is another area where the elation of having developed a successful product or service can quickly give way to sticker shock and scrambles when the next cloud bills arrive. When the cloud is literally your means of delivery, keeping costs corralled can seem an impossible challenge, at odds with growth — and even mere survival.
[RELATED: Exploring Career Opportunities in Cloud Computing.]
Zenedge, a company that was founded to help secure critical IT systems, experienced this exact challenge when expanding to manage these systems via the cloud. Within three years, in fact, the company’s cloud services bills had ballooned from a mere thousand dollars to a whopping $1.2 million — every month! Founder Laurent Gil elaborated, “… [It was] the life of a SaaS product. We were using our cloud to deliver our product, so the more customers we had, the more costs we had. That’s normal… The frustration was not being able to understand what to do about it.”
A frustration that Gil, and his Zenedge cofounders, didn’t tolerate for long: Realizing that other companies were facing down the same struggle, they rose to the challenge. Their new company, called CAST AI, helps businesses automatically optimize their Kubernetes clusters in real-time. Since usage fluctuates, CAST AI’s engine is also trained to reconfigure and reoptimize every few seconds for a la minute rightsizing. In the first 14 months after CAST AI’s launch, this functionality saved customers between 50-75% on their cloud expenditures, with an average of around 63%.
AI for Real-Time Network Monitoring
Operations that have to take place in real-time, or close to it, are another area where AI can make an immediate and very visible impact for organizations. A large UK-based mobile phone carrier needed a way to monitor several of their web-based services against any deviations in performance. As a first step, they approached Torry Harris Integration Solutions (THIS), a British IT consultancy firm specializing in software strategies that support digital business transformations, to develop more robust network monitoring capabilities. In response, THIS’ strategy involved solutions modeled out using machine learning to scan for glitches and errors, understanding that relying on humans would hardly be sufficient. Now that the mobile company is able to essentially monitor around the clock, network anomalies are flagged much earlier, which helps to get them resolved before they wind up causing any server strain or downtime.
AI and Predictive Analytics
AI and ML-derived tech is already helping companies everywhere make strong plans to move forward into an increasingly data-driven future. “AI and predictive analytics [are] permeating your organization and you’re not even necessarily aware of it,” according to David Kuder (Head of Cognitive Insights and Engagement, Deloitte). Most commonly, this is done via predictive algorithmic forecasting, which is a process by which AI is provided with historical/empirical data sets and previous predictions and then prompted for insight into current strategy, as Cognizant recently showcased when implementing their SAP Analytics Cloud solution for Elizabeth River Crossings (ERC), a VA-based transportation infrastructure company.
Initially seeking more actionable BI insights to help manage the Elizabeth River Tunnels Project, which spans the route between Portsmouth and Norfolk, ERC discovered they now had a 360-degree perspective of the business. With more capacity to access and combine data from different sources, custom reports, and visualizations take minutes to achieve. And since even very intensive data services jobs can now be run in less than an hour, and not seven hours, the only question left for ERC is what they might do with all of that free time.
Conclusion
Properly managing a cloud program is a challenging task and one that becomes harder by the day. Serendipitously, perhaps, it’s fitting that just as we conclude it’s too big a job for one person (or one whole team), technological innovation has produced usable AI. Now that we humans can delegate many of the challenges of the cloud—like automation, migrations, cloud cost management, actionable insights, and more — we’re free to take much bigger steps toward unlocking its full value.
