We often discuss the many ways in which IPsoft’s automation solutions can be utilized as self-healing tools for IT operations. Rather than react to changes in an environment, IT teams can use 1Desk, the autonomic backbone for the enterprise, and IPcenter, the groundbreaking automated IT operations platform, to automatically monitor and resolve problems before they become major issues. Enabling this kind of capability was the focus of a conversation led by Andrew Belzile, Program Manager at CGI, during the annual Digital Workforce Summit.
CGI is a global IT and business consulting firm based in Montreal. The company provides clients with systems integration, managed services and proprietary IT solutions to help run businesses more efficiently. CGI has been an IPcenter client for the past four years.
During the past 12 months, CGI has been using a custom automation within IPcenter that allows the solution to function as a cognitive predictive analytics engine, which enables better resource utilization and improved cost-effectiveness, Belzile said.
“Using data that we’ve been collecting with IPcenter over the last four years, one of our very brilliant engineers developed an algorithm that can help us predict storage usage for our customers,” said Belzile. “You can see spikes of servers go up maybe once or twice a year at 20%, then go straight back down. So instead of having our customers pay for storage at maximum capacity all year long, we can offer more flexible rates to them.”
In essence, the engineer customized IPcenter to predict, on its own, how much storage CGI needs across its entire roster of customers, as well as its own internal data center.
Additionally, the company applied the algorithm to its ELK Stack to collect log data. The algorithm isn’t yet live for CGI clients, but Belzile said he hopes to make it available “shortly, at least in a beta phase” to combine collected log data with IPsoft’s monitoring and IT service management data.
“Using data that we’ve been collecting with IPcenter over the last four years, one of our very brilliant engineers developed an algorithm that can help us predict storage usage for our customers.”
— Andrew Belzile, Program Manager, CGI
In a specific example of how this algorithm would be applied to a live IT environment, Belzile said a Priority 1 or Priority 2 ticket would trigger the automation to correlate the buildup of logs on those servers or specific applications. “Sometimes it can give us 5 minutes, 10 minutes, 15 minutes, an hour of advanced notice of something coming up so that we can react before there’s actual customer impact,” he said.
To read additional commentary from DWS 2019, be sure to review this post about Telefoníca’s Amelia implementation. You can also read this article from Information Age about how companies can achieve ROI by avoiding basic AI implementation mistakes. For additional DWS perspectives including media coverage, head over to our IPsoft Newsroom.