Digital Business

Separating the signal from the noise with AI-assisted user experience monitoring

Dec 05, 2023
ZDX and AI

Throughout 2024, organizations will continue needing help creating a great end-user digital experience all while increasingly adopting digital-first business practices. The help desk, desktop support, and network teams bear the brunt of the frustration. Users from across the organization come to them because they can't get something done or something's broken and they expect fast remediation. These teams may be a step removed from end users, but they play a decisive role in enabling businesses to succeed.

The consensus in the industry is that the tools for end-user device monitoring, network monitoring, and application monitoring are fragmented and siloed because they were all used by separate functional teams. The security team also feels the pain since network devices like firewalls are usually under the networking team's purview. When these groups use their tools to troubleshoot issues, they each have different combinations and capabilities that add to the challenge. 

"I was once with the financial organization, and they did a tools rationalization, and they came up with 16 tools, and I said to the CTO, 'Wow, I am shocked you only have 16 tools,'" recalls Zscaler VP of Product Marketing Krishnan Badrinarayanan and expert in digital monitoring solutions. "But the CTO viewed 16 as too many and inefficient."  

The CTO may be right. We have to change mindsets to abolish the status quo and radically simplify how we maintain productive and happy users. 

So, how can we do things differently? 

The answer starts with zero trust architecture, which changes everything about access and, in turn, transforms how you monitor user experience. Network monitoring tools are headed toward extinction in the zero-trust world since IT operations will get the insights they need from the endpoint right up to the application.

According to Badrinarayanan, Zscaler is giving customers the visibility they need to be successful, starting with the consolidation of telemetry across endpoints, networks, and applications, as well as zero trust architectures that sit in between and other cloud proxies that you might have in place. One unified place pulls all the data and insights for security, network, and service desk teams. 

Before we get into the details, remember that it all starts with the data. You need a consolidated way to collect telemetry across the entire service delivery chain from the end user's device across networks, through cloud proxies, firewalls, or anything else right up to the application, irrespective of where the application lives, whether it's a SaaS-based application sitting in a private data center or a private app that's sitting on infrastructure as a service from a cloud service provider. AI and ML then really span the distance to make sense of that data in order to provide answers versus hundreds more data points to you.

Specifically, AI correlates all the data gathered for an incident from the end user device across the network to the app. The machine learning model behind the AI can pinpoint the problem for the service desk analyst. If It's a local ISP issue, the service desk team can escalate that precisely to the network team to have them resolve it. If it's a WiFi issue, such as the end user using a 2.4 GHz vs. 5 GHz band, the service desk team can recommend switching bands, and he or she can get back to work.

This scenario is much better than an end user raising a ticket that gets flagged across 15 groups, and hopefully, one of them will raise their hand and say, "Yes, this is where the problem is." Think about the wasted effort and time across IT resources, not to mention the frustration and weeks of delay the end user faces.

That's why AI is essential, and that's just a service desk team. The same holds true for network operations teams, who can use these models to gather insights on developing issues impacting specific offices, regions, and departments and get ahead of developing issues instead of reacting to problems already occurring. 

The service desk security or the network teams are now able to collaborate effectively and respond quickly. With Zscaler Digital Experience or ZDX (a multi-tenant, cloud-based monitoring platform to probe, benchmark, and measure the digital experiences for every single user within your organization), a service desk team member can pick up a ticket and isolate the issue. If it is a network issue, the service desk employee can easily share that information through an escalation with the network team and the network team. If that team deems the issue stems from a firewall, they can then contact security. Teams come together in a significant way using insights and answers versus cryptic data that exists across various silos.

Customers experience mean time to resolve (MTTR) that were once weeks or months are whittled down to hours or minutes. Zscaler offers customers the visibility into zero trust environments using AI combined with data from all the points within the service delivery chain. That is just the tip of the iceberg. Layered on top of the AI-based root cause analysis baked into ZDX is the ability to intelligently identify anomalies and automatically create support tickets in your service management tool. A service desk analyst can quickly route those tickets to the right teams or solve them themselves. 

Moreover, end users can solve some problems themselves. The Zscaler Client Connector has an AI engine that constantly monitors the performance of your end-user device (in addition to providing security) and makes recommendations to improve speed and quality, such as by freeing up CPU cycles. In turn, service desk teams can deflect a considerable number of tickets.

Another key capability for networking teams is incident dashboards. Incidents are issues that impact the device performance of multiple users, and the dashboard displays them across four types: WiFi, Last Mile ISP, Zscaler Data Center, and Application. AI/ML detects and identifies incidents using the appropriate metrics correlating to the issues. The incidents displayed are based on the selected time frame range in the UI and show incidents over time impacted users, and where they occur on the map.  

“We process more traffic than some of the biggest companies in the world through the Zscaler Zero Trust Exchange, which puts us in a unique position to offer our customers a crystal ball," said Badrinarayanan. 

What to read next 

The employee experience is digital and measurable: how do you score?

Zero trust and private app visibility: a new ray of light for digital experience