Use Of AI To Pinpoint Problems

By Karen Falcone, Juniper Networks

Troubleshooting customer problems can be one of the biggest wastes of time for an MSP. It’s hard being judged by how fast service can be restored - whether the company was the cause of the problem or not. Every second of downtime increases their dissatisfaction.

Customers care about one thing and one thing alone - whether the network works. Their expectations are for near-constant uptime, and when that isn’t the case, the MSP is often the first to blame. As networking companies and service providers across the industry look to redefine networking success through the lens of the end user experience, those expectations will now be front and center in any contracts and service requirements.

With that in mind, MSPs will need to take extra care to ensure they’re providing a positive end user experience, at all levels, or there will be a great deal of churn on the horizon.

Focus On The Experience

End-users remember when services go down. It’s easy to forget the thousands of times they connected to the network and downloaded documents with ease when a video call with an important client drops midway through.

Key to providing end users with a positive experience will be identifying, understanding, and solving problems quickly and efficiently. That, however, is often easier said than done. Over the past few years, there has been a massive increase in the number of devices, applications, bandwidth, and more that an MSP has to deal with for even a single customer. In addition, since the pandemic-driven rise of hybrid and remote working, these instances can now be multiplied many times over. End-users are using a mix of corporate and personal resources from several different locations and devices, all with their own specific variables that need to be considered.

With organizations’ 24/7 reliance on cloud-based applications, the attack surface, as well as the time and effort it takes to maintain the network environment have increased exponentially. This requires constant attention by a skilled team that can quickly recognize and repair any operational or security issues.

Understanding and maintaining the modern network infrastructure is tough enough for an MSP to successfully handle in these current conditions, before even factoring in what to do when something goes wrong and it needs to be rapidly repaired.

Applying Artificial Intelligence

Traditionally, when problems in the network arose, the MSP would then need to begin a lengthy troubleshooting process to narrow down the cause of the issue. It was barely humanly possible to quickly track down and correct an issue without some luck - and that was when most MSPs would simply need to visit a corporate office in person and troubleshoot for a few hours.

This is where Artificial Intelligence (AI) can help. By applying AI to network management, problems can be pinpointed faster, and an organization can even receive assistance in fixing it. AI can immediately make it easier to uncover problems such as wireless network interference issues, network misconfigurations, incorrect settings, missing Virtual LANs, or even bad cables. And faster information and quicker fixes are what will protect the positive end user experience that MSPs are expected to provide.

There are a few key ways that AI can help shorten troubleshooting time and speed repairs:

• Locating Issues - Say, for example, that an organization is experiencing a dead zone in a certain area of their office space. The employees who are located in that area complain of being unable to connect, or of being unable to download and access documents as fast as they can in other parts of the office. Clearly, there’s an issue. In the past, an MSP’s technician would need to wander around the office, using trial-and-error to determine the area of the outage. Following that, they’d need to then troubleshoot what could be the issue. Computer software not updated? Bad router? Misconfiguration to blame? Bad cables?

Trying out each possibility would take time and only lengthen employee dissatisfaction. By applying AI to the issue, both the area of the outage and what’s causing it could be easily identified, saving time and getting end users back up and running faster.

• Alert Management - Be it operational or security, enterprise monitoring and defense tools review the network environment and provide alerts to the IT/network management/security teams, letting them know of potential issues. In most cases, the settings on these tools are configured in a way that more alerts are sent, so that potential issues are not missed. However, when the surface expands the way it has recently, it’s easy to become overwhelmed with alerts.

AI can help an MSP weed through these alerts and others to locate specific issues that should be addressed and eliminate false positives. That way, an MSP technician can spend their time on what matters most to the end user experience.

• Recommending Repairs - AI isn’t just for finding issues, however. Another powerful way that it can help an MSP speed repairs and get end users back to work is by helping technicians make the actual repairs. In many cases, finding an issue is only the first step. Once found, the technician needs to solve it. Sometimes training and experience can still come up short, leaving a technician to call in a colleague or look to others (partners, vendors, etc.) to help understand what needs to be fixed. With AI tools in place, however, repair steps and more can be provided directly to the technician, helping correct the issue sooner and without the need to call in a team of experts.
   
• Becoming Proactive - Continually operating from a reactive stance, answering complaints, and providing fixes, can be tiring to an MSP’s team and give off the false impression of not doing one’s job correctly. Improving the ability to be proactive is another way that AI can help. For example, when it comes to wireless networks, it is common knowledge that there are a number of problems that users ignore or forget. These problems can pile up and eventually affect the end user experience. AI can learn to spot issues before they become complaints, allowing for proactive correction before an end user was ever aware of an issue.

Helping The MSP Succeed

In the current work environment, businesses of all types are under pressure to do more with less. MSPs are pressured to provide more and more services for their customers, regardless of how much more complex the customer environment has become. This is another way that AI can help. With AI-driven tools in place, MSPs can help their technicians work more efficiently and eliminate mundane tasks such as testing each computer to pinpoint a misconfiguration. By streamlining these tasks and making their technicians’ lives easier, MSPs will be able to better retain the talent they have, and be able to provide more proactive, business-critical services to customers.

AI is truly the key to a better end user experience for all.

About The Author

Karen Falcone serves as Sr. Director of Product Marketing for AI-driven SD-WAN, and Service Provider as a Channel at Juniper Networks. Prior to this role, Karen served as Vice President of Marketing at 128 Technology, which was acquired by Juniper in 2020. Karen is enjoying her tenure at Juniper Networks, having been a member of the Unisphere Networks team acquired by Juniper in 2002. Karen holds a BA in Finance and an MBA in Marketing from Bentley University.