AI: Stress Relief for IT’s “New Normal”

In this special guest feature, Rachel Obstler, VP of Product at PagerDuty, focuses on the power for data analytics and AI during today’s current environment. Rachel is responsible for the product direction, customer experience, and pricing. Prior to PagerDuty, she served as the VP of Product at Keynote Systems, and later as the VP of Mobile Testing, overseeing the mobile testing product, sales, engineering, marketing and service organizations. Rachel has over 10 years of experience in SaaS and over 15 years in product management. She holds a B.S. at MIT and M.B.A at Stanford University Graduate School of Business.

Pressure has long been the norm for IT incident response teams, but shelter at home mandates ramped up the stress levels in IT organizations beyond anything anybody ever imagined. Suddenly, IT teams weren’t just working hard to keep services running perfectly — they were designing new digital services that cater to increases in online traffic and demand, helping the business pivot to support remote work-forces, and spinning up their own crisis response teams, all while working in a new remote environment themselves.

With over 80 percent of office employees working from home, many digital services saw surging demand, which caused an unprecedented strain on IT systems and teams. In March, incident response teams on average were dealing with at least double their normal workload — with verticals like online learning experiencing an 11x increase in incidents, and collaboration services not far behind with an increase of 8.5x. 

Problems that generate incidents may range from lack of real-time updates for a delivery app to slowdowns in a search function to delays during a video conference. Whatever the issue is, there’s a new sense of urgency now that for many organizations, digital business is their only business. As a result, many IT teams are operating in a prolonged period of “hypercare,” or an elevated level of support typically associated with major changes or high traffic events like Black Friday. 

AI is a quick study

Successfully managing an onslaught of incidents that comes with heightened demand ultimately depends on access to actionable intelligence in real-time. These incident response teams are putting data to work faster and more effectively than ever before and we’re seeing this level of success today in our customer base in spite of the increased workload with teams resolving issues up to 60 percent faster than pre-pandemic. It’s critical to automate processes to create efficiencies — nothing is more important than getting the right information to the right people at the right times. AI contributes to this success in several ways.

• Noise reduction and intelligent triage. The huge amount of data now available can present itself as chaotic and distracting noise, actually making it more difficult to know which issues need to be addressed immediately and how to address them effectively. AI can assess, filter and organize the flow of data associated with incidents and manage the allocation of resources based on priority. Some types of incidents, for example, can temporarily be put on hold to free resources for more urgent problems. Another way to avoid overloading first response teams is to group related alerts into a single problem which eliminates duplicative efforts and cuts down resolution time. 

• Intelligent deployment of resources. AI can suggest appropriate individuals or teams to assign issues to, based not only on scheduling and availability but also on the nature of the incident itself. By doing so, response teams are empowered to fix issues faster, reducing costs and downtime, and protecting the customer relationship. 

• Faster remediation. AI can connect current incidents with similar incidents from the past, indicate who worked on them and what remediation steps were taken. It can also uncover links between different teams’ incidents that seem unrelated and show that they’re all part of the same problem. All of these capabilities lead to faster incident resolution while eliminating the duplication of effort that occurs when teams are siloed. 

Using AI to keep customers happy

51% of companies find out about issues with their digital services from their customers, which can be catastrophic when your business relies on digital interactions and delivering stellar customer experience. To combat this, leading organizations focus on identifying and resolving issues before their customers notice anything is wrong. Sophisticated AI and data analytics make this goal achievable. 

One of the most important benefits of AI-based systems and processes is that they don’t stop learning once they’re deployed. Instead, they get better at what they do – and this aligns perfectly with the long-term goals of companies that depend on trustworthy online services to maintain customer loyalty. 

Sign up for the free insideBIGDATA newsletter.