Why Aurora AI Agents Only Reduce Ticket Load When Built on Accurate Network Data — Steve Petryschuk

[00:00:01] Dave Sobel here on the floor of ChannelCon, sitting with Steve Petryschuk. He is the VP of Product and Market Strategy at Auvik. Now back in March, Auvik published its 2026 IT Trends Report and at the end of April launched Aurora, its IT agent suite for IT operations. The report found that 67% of IT pros are optimistic about AI and 5% say AI is actually core to how they work every day. And that gap is what we're going to talk about. Steve, welcome to the show. Steve Petryschuk Thank you so much for having me, Dave. It's great to be here.

[00:00:30] Dave Well, I'm excited to have you. So let's start with the Aurora launch in April, right? Give me the update in your words, like what shipped, what's in customers' hands, and what changed? Steve Petryschuk Yeah, so Auvik launched our Auvik AI, branded as Aurora agents that are helping IT teams operationalize artificial intelligence within network operations. And so going beyond just slapping a chat bot on top of their existing technology, these Auvik agents use network context built on the foundations of Auvik years of collecting network data, and are specialized

[00:01:04] Steve Petryschuk So it's like the headline feature. There's got to be something in that pile that people are using right now and are excited about. Steve Petryschuk My favorite one is the Aurora troubleshooting assistance. It's built into every single alert that Auvik generates. Oftentimes when MSPs get an alert in their inbox, the tier one tech, who may not be the network expert, right? They're the more generalists. They haven't gone to the

[00:01:28] school to be a CCIE trained expert. They don't necessarily know what to do next, how to tackle the problem, what the root cause might be. And so these Aurora troubleshooting agents help them with context, tell them exactly taking context about connectivity, and network performance issues, and syslog data and all these technical things to help them troubleshoot that issue more quickly and get the end user back on the road. It's had a ton of feedback and even some feedback here on the channel con show floor from a lot of our longtime partners, basically saying, finally, a real

[00:01:58] applicable, valuable use case of AI. Steve Petryschuk So Auvik's added server endpoint and SaaS management monitor over the last two years. What were MSPs running into that made that expansion become necessary rather than optional? Steve Petryschuk If we start off with the endpoint side, getting more visibility into that complete end user experience from the device all the way to the application that the user's doing to do their job, our SaaS monitoring and endpoint monitoring give you that visibility into how the network is performing, what applications users might be using.

[00:02:28] Steve Petryschuk So when I go to troubleshoot these network issues for users, I have all the context I need to be able to solve those problems. And increasingly, when we talk about the server monitoring capabilities, I'm looking for more visibility into the network side of the performance on all my servers and network assets within the same dashboard. Steve Petryschuk So it's largely been a consolidation of tools, maybe taking some data that they may have had in a separate network monitoring platform before and being able to simplify their tool stack with fewer tools.

[00:02:55] Steve Petryschuk Now, if I pull from that report, 67% of IT pros are optimistic about AI. 5% say it's core to their daily work. So you're not shipping AI agents into the 5%, essentially the ones that are core. What has to be true for that number to move? Like, do you view that as a product problem or a customer readiness problem? Steve Petryschuk Steve Petryschuk So when we did the analysis and surveyed our customers, what we found is that a lot of MSPs, that was 67% are really driving some experimentation and initial adoption of AI.

[00:03:23] Steve Petryschuk But I think for AI to really be core to their network operations, to see that 5% increase to something more significant, we have to see repeatable value from the AI use cases beyond the experimentation. Steve Petryschuk What we see a lot with the experimentation is that the IT technicians might, you know, have the AI get them 50% of the way there and then it's their expertise that helps finish the job.

[00:03:44] Steve Petryschuk We need to have AI getting them to 90%, 95% of that problem resolved for them to really establish trust within the AI systems and start to maybe not hand over the keys to resolving problems or to full issue resolution, but to be able to at least trust the AI to troubleshoot problems for them. Steve Petryschuk And so what we focused at Auvik is building a lot of troubleshooting agents on top of our core expertise, right?

[00:04:09] Steve Petryschuk That is the network monitoring, network management expertise, so that we can build out efficient AI workflows and make these AI workflows part of the core operations. Steve Petryschuk This episode is supported by ScalePad. Steve Petryschuk Here's the uncomfortable part of running a good MSP. Steve Petryschuk The client decides whether to keep you based on what they can see, and most of what you do well is invisible by design. Steve Petryschuk Nothing broke, nobody noticed.

[00:04:36] ScalePad's Life Cycle Manager is built for that gap, bringing the full picture for each client together, so you can show how your work connects to their business goals. Steve Petryschuk The work may be technical, but the value should be obvious to the person paying for it. ScalePad.com slash MSP Radio Steve Petryschuk Now, I found out interesting, you told Channel Buzz that the first thing MSPs need is documentation.

[00:05:02] An up-to-date source of truth is what you said, because that's what the AI has to build on. Most MSPs know their documentation isn't that great. They know this going in. So, are you selling AI agents into a market that you've kind of said isn't ready for them? Steve Petryschuk I don't think that's the case. I think if you look at the general landscape of where Auvik provides value, our first mission is to go out and discover and document the entire IT landscape. Identify every asset that's connected into the network.

[00:05:30] Identify all the applications that users are using to get their job done. So, I think it aligns with Auvik's core value prop. I do believe that there's still challenges in having up-to-date documentation and ensuring that when you start to build AI workflows, that it's on a solid foundation of authoritative knowledge on what's out there, what assets I have to manage. But that's part of the problem that Auvik helps solve. And so, we're excited to be part of the documentation solution and then part of the building the automations with AI on top.

[00:05:56] So, I think based on that, you've got criteria now for what separates an AI agent that reduces ticket load from one that just adds a dashboard. Like, give me the versions of what the differences that MSPs can measure on the ones that are successful. So, when we look at successful agents versus unsuccessful AI agents, I look towards does it add revenue or save costs? Or does it drive overall efficiency within my business? And so, I'll sort of fuck it. There's two different things that drive revenue or save costs.

[00:06:27] And the Aurora agents can do both of those. We have Auvik Aurora agents that help build out a full hardware lifecycle visibility. And so, as we discover new assets, these Aurora agents behind the scenes are collecting data about end-of-life, end-of-support dates to start to build out technology roadmap plans and to give insights into hardware refresh opportunities for MSPs, thereby giving them these revenue opportunities. So, we've proven value down that use case.

[00:06:51] When we look at things like the Aurora troubleshooting agents, where we're getting tickets into the inbox and we're helping IT technicians solve those quicker, that's now adding efficiency and saving them costs through their journey. And so, both of these things are going to help build a demonstrable use case where Auvik Aurora AI agents can drive more efficiency into business. Interestingly, Auvik detected more than 100,000 shadow AI applications on customer networks in 2025.

[00:07:16] So, walk me through what that number means practically for an MSP, say, running 40 clients. Like, what does that look like in practice? When you look at the number of shadow AI applications in a typical customer environment, you're talking, you know, tens, dozens, sometimes up into 100, depending on the size of that end client. There's sort of a proportional relationship to the number of employees within a specific client and the number of shadow AI applications. I think what's obvious across the board is that our customers are interested in AI.

[00:07:46] They are starting to adopt AI. And whether they have a formal AI program or not, it is making its way into their environment. And so, these end clients are looking for support, looking for guidance on what types of safe AI use will look like. And so, it's on us as service providers to start to shine some light on what the scope of that problem is, identifying those shadow AI applications for them, and then helping them build a roadmap or building a plan to help control and enforce and govern those AI applications.

[00:08:15] Not eliminate them, but understand what the safe usage of those are, and then add some guardrails to make sure that we're using it in a safe, efficient way, and using the approved tools that we should be using. Now, your research also says about 60% of IT teams are finding unauthorized SaaS at least monthly. So, when does this discovery stop being a report you work with of a client and start becoming an enforcement problem that the MSP owns? That's a great question.

[00:08:41] When we look at this becoming an enforcement problem, I don't think that this is an enforcement that the MSP can take on their own. Obviously, there are a lot of downstream implications to shadow AI usage and shadow SaaS usage in terms of security controls and additional paths for threats to enter the environment. But we can't unilaterally enforce a policy on our clients that restricts the use of these shadow SaaS tools.

[00:09:04] It's about first educating, creating awareness, sharing through the, whether it's through a QBR process or a customer business review where I'm sharing the shadow AI applications entering the environment. I can start to shine some light on those and start to inform the clients about the risks that they pose and build with them a collaborative plan to help bring those under control. I don't think the reporting ever goes away. We'll continue to always have shadow SaaS applications. It's just a fact of life these days.

[00:09:31] But ensuring that it becomes part of our standard operating procedure to identify what those are, highlight the risk of those, and add in safe usage of those SaaS applications. That's the repeatable model that we'll move to. Are you and your clients tired of the time-consuming ticket tennis of coordinating meetings and help desk calls? Wouldn't it be better to automate this process with a tool that connects directly to ConnectWise Manage or Autotask?

[00:09:55] TimeZest offers scheduling automation that gives you complete control of your schedule and eliminates the hassle of calendar ping pong. As the only service designed specifically for MSPs, it integrates into your workflow and makes scheduling appointments easy on you and your clients. Plus, you can try TimeZest for free.

[00:10:16] Visit TimeZest.com slash MSPRadio and use the code MSPRadio to get 10% off your first year of TimeZest. Now, I want to ask you a couple of questions about what the community conversation is. So I'm going to be trying to be citing exactly where my numbers come from so I can be as fair as possible about it. Sure. So your own research says that IT teams are time-crunched, understaffed, and governance deficient, right? And Auvik's answer there is more software.

[00:10:44] So make the case that Aurora subtracts from an overwhelmed team rather than adding to it. And can you give me a number that you put to that? That's a great question. So when we look down at how Auvik can help add efficiency or improve the lives of the average technician, I'll actually look back to a story that I heard just two weeks ago from one of our newer users of the Auvik Aurora troubleshooting agents.

[00:11:08] This user had gotten an alert from Auvik that had produced a high storage utilization alert. They were a tier one technician who aren't super familiar with network devices, but they were able to use the troubleshooting assistance to identify that this alert is often triggered by items like old configuration files that maybe need to be cleaned up. And it also described what the impacts of not resolving this alert might be, like the switch becoming unresponsive or the switch creating downtime.

[00:11:36] And so by quickly understanding the impact and the possible causes, this tier one technician was able to go and resolve that issue. Now, typically, the old model would be that alert has to go to a tier three tech because the tier one doesn't have the context. So they've now spent time on that ticket. They've escalated to someone else. And now we've lost efficiency. Now we have two people looking at the same thing. It's going to take longer to resolve. So in this case, Auvik Aurora is accelerating the troubleshooting by giving more tools to the people who are seeing that ticket first.

[00:12:04] And so if I was to put a number on how much we're reducing that troubleshooting time, I would say it's about 50%. Obviously, your miles will vary depending on how many network type issues you have. But being able to produce any troubleshooting time for any network issues is super valuable. So I want to give you credit where it's due. In our community tracking, people consistently say that Auvik has the best mapping and best user interface in your category. But after the price increase, the same threads are also talking about some migrations. They're looking at cheaper alternatives.

[00:12:32] So what does an MSP get? So when you look at the landscape for network mining systems, I believe that putting a healthy, competitive landscape is actually unofficial to all MSPs and the whole channel. It drives innovation. It drives accountability to organizations like Auvik to ensure that our solutions continue to evolve and continue to provide a ton of value to our clients. And that the item that you brought up, our core network mapping capability, is still our core differentiator. And it's something that, to be honest, no one in the industry has been able to catch up to yet in that bid market space.

[00:13:02] So if I was to look in the future and say, you know, why continue to choose Auvik? It's all about the roadmap and the items that we're continuing to deliver on. We have a vision of a future where, you know, network monitoring is autonomous. Network change is autonomous. And so we're continuing to build out more infrastructure around pushing configuration changes to network devices, more ability to monitor cloud infrastructure. And it's that continued innovation, combined with all the innovations we're making within AI, to give MSPs a framework to start to scale network monitoring and network management and this is autonomous future.

[00:13:31] So I'm going to ask you to make the case against your product for me for a moment. I'm going to say most MSPs really can't act on all the alerts they have, right? They've got all this mess. Like, why is more visibility the right answer? When you're looking across a network infrastructure and getting a signal in form of an alert, that is telling you something. Saying that I have a ton of alerts and there's too many of them is actually sort of like burying your head in the sand and saying, you know what, I'm just getting too much. I'm not going to pay attention to it anymore.

[00:13:57] Now, understandably, it's a real challenge to tackle every single alert, to resolve every single alert. So at all of it, we've actually leveraged our Aurora Aegeus to start to build out our alert insights report that analyzes these alerts that trigger very often and tell you and start to correlate some of those alerts and look for that signal in the noise. Oh, these alerts are happening often because of this specific network issue that may have gone unnoticed before. Or you're getting too many of these type of alerts, but if you change the configuration of that alert slightly, every signal becomes more valuable.

[00:14:26] So while I would never recommend to ignore the alerts, ultimately, I want every alert that lands in someone's inbox. It's actionable and relevant. So as we wrap up our time here, what trend do you think is going to have the biggest operational impact on MSPs in the next 18 months? And what do you think they should be doing about it this quarter? But if I look towards the 18 months from now, I think we have to move from a spot where we have independent AI agents working within our different tool silos to a more integrated approach.

[00:14:54] And so MSPs should be redoing their tool stack today to understand what each one of their vendors is doing to help provide their insights and their niche knowledge to the other players in the space. Auvik's unique knowledge is within the network monitoring, network management space. We are the experts there. Leverage our agents, leverage our insights to help inform the rest of your tool stack. And ask vendors like us, what are we doing to help you sort of support this autonomous future leveraging our skill set, our expertise? Steve, I've learned a ton. I really appreciate you joining me today. Thank you very much for having me.

[00:15:24] The hardest part of running an MSP? Doing it alone. The Small Biz Thoughts community has been the room where independent operators compare notes for nearly 20 years. Real peers, real numbers, real answers from people running businesses just like yours. Pull up a chair at smallbizthoughts.org. Interested in advertising? Head to mspradio.com slash engage.

[00:15:51] The Business of Tech is written and produced by me, Dave Sobel, under ethics guidelines posted at businessof.tech. Thanks for listening. I'll see you on the next episode. Produced by Picture This Video. Part of the MSP Radio Network.