Transcript
You're going to see autopilot in action, and we're going to have a good time. We'll go over the purpose of the feature as well as the demo, and we'll talk about what makes autopilot unique. We have myself, Alina, as your customer success manager here at Terra, and we have Uri, the account executive, who's going to be showing us all of the ins and outs of the autopilot. A little bit of a housekeeping before we proceed. So this session is being recorded, and it's going to be shared with you after the webinar. We also invite you to type in questions in the Q&A window. You have a chat window as well as the Q&A. So if you have any questions, feel free to drop them in the Q&A section. At the end of this webinar, we're going to ask you to fill in a short story. It's going to take less than a minute. We would love to hear your response. Now, you're going to see a poll pop on your screen. Please let us know if currently you're using AI in your day-to-day. There is no right or wrong question. We just want to understand how comfortable, I guess, you are with AI, how interested you are in this topic. We're getting some answers. Okay. That's good. Most of the answers we're coming in, we see that you cannot live without it, which is great. It means you're doing the right webinar today. Let's give it a few seconds to gather all of the answers, right? And now, you're also going to have another poll, which is just to also let us know which one of these tasks take most of your time currently. Are they resetting passports, handling repetitive and manual tier one support tasks, managing too many different IT tools, or user frustration and poor end-user experience? All right. We have the answers coming in. Some are handling repetitive tasks. Some are managing too many different IT tools. We also have some user frustration. All right. Thank you so much for taking the time to answer the poll. Uri, I'm going to stop sharing my screen, and the stage is yours. Thanks so much. Let me share my screen. Okay. Just waiting for that to come up. There we go. Hi. Thanks so much, Alina. So, hi again, everyone. Today, as Alina already introed to us, we're going to learn a little bit more about what autonomous IT actually is, and how Atera is sort of leading this space to go ahead and help yourselves in your day-to-day. Before we actually present what autonomous IT is all about, which we'll be seeing today in a demo, we're actually going to first talk about a bit of the challenge before we move into the actual demo itself. So, as you can see on my screen here, these are just some different stats that we've collected through different type of research that's been done out in the field, and really it boils down to on the right-hand side that annual IT downtime costs companies a lot of money. Okay? Whether this is due to different types of shortages on the IT side for talent, or just because people, you know, with so many, really a plethora, so many different types of systems they use day in, day out, people do get confused, leading to an increase in tickets. Okay? Atera has tackled this space with not one, but two different tools in the AI space. On the left-hand side, the AI Copilot is really built for the technicians. This is to give the technicians inside of Atera different types of tools in order to increase their efficiency on the day-to-day. We have a different webinar, which we'll be happy to share with the audience here, or you can find on our website, which goes into different capabilities of the AI Copilot. Today, we'll be focusing on the IT Autopilot on the right-hand side. For those of you who have used the AI Copilot, Atera has built a solution for end users that builds on some of the capabilities we've released with the AI Copilot already as far back as almost three years. The IT Autopilot now makes this available for end users in order to go ahead and just really let yourselves work, right? At the end of the day, you all are getting to your desks at IT and trying to get stuff done, and you're being pulled in either three, five, or perhaps 17 directions, and that's only before lunch. We can allow you to take the focus for these, especially Tier 1, but even Tier 2 and Tier 3 types of issues, and offload it to the AI to do it for you. The value this really brings, well, it's as you can see here from a couple of different areas. For the end users, there's no more of that ping pong, no more of that waiting around for perhaps the IT team to respond or whatever the case is. Rather, they're able to get responses in real time. For IT teams, again, by being able to offload a lot of this to the AI, you're able to free up your time in order to go ahead and focus on higher priority things. Even in the event that the AI is unable to autonomously fully resolve the ticket, it will still save the IT team's time because, in essence, the AI will have already performed those first couple of actions on its own, so you can skip ahead in your troubleshooting slash diagnostic process. And then, of course, for the organization, as we reduce IT downtime across the org, this only brings value to the organization by time saved and time at the end of the day is money. Now, what makes the IT Autopilot unique is the fact that it is the only native AI agent specifically crafted for autonomous IT. And really, the uniqueness about it is because it's paired with the Terra's RMM agent, it's able to perform not just things like surfacing knowledge-based articles, but it's able to also go ahead and perform what's called device actions. And by leveraging those capabilities, and we're going to see some of these examples today, it's able to fully autonomously resolve different types of use cases. Okay, let's dive into a demo and walk through this process. Okay. Now, one thing that is important to note is that the Terra, the IT Autopilot is an omni-channel type of engagement. While that could be through your ticketing portal, for those of you who leverage through a chat experience, this could also be through Slack, Teams, it could even be through email. So imagine you're using email today, instead of that 30 minutes, hour until the user gets a response, they get a response near instantaneously. As I mentioned, the IT Autopilot has three different values it can bring to the organization. The first we mentioned, device actions, where by pairing with the RMM agent, it's able to go ahead and perform an action autonomously. The second I had also mentioned, which is being able to surface information in the nick of time, in real time. So KB articles, knowledge-based articles, which you have created for your organization. The third is also cloud actions. Cloud actions allow you to connect separate third-party systems through a myriad of different ways. That could be via API, MCP, Zapier, for those of you who are perhaps familiar with, in order to go ahead and leverage third-party use cases. That could be something like opening a task on a Monday board, taking a vacation in the vacation system, or whatever that might be. Today we will showcase for you two different types of use cases. The first one that I'm going to show you, I'm going to bring this here to my other screen if you'll bear with me a moment, is actually a surfacing knowledge base. So this is my actual Slack, because that's what I have for now. And you'll see that as an end user here in Eterra, I have an issue. My issue is, is that I have a guest here and I am unfamiliar with the Wi-Fi guest password. So instead of waiting around for my great IT colleagues to go ahead and give it to me, I can simply leverage the autopilot in the right-hand side of my Slack to show me what is the password for the guest Wi-Fi, okay? Guaranteed that if I had actually raised this ticket as a real ticket, I'm probably on the bottom of that priority totem pole, if you will, right? Because they have better things to do than just give me this Wi-Fi password. But I, as an end user, don't have to wait around. This will scan all of the knowledge base articles made public by the IT team and has given me that answer in real time. I can now move on to the next piece of my day, all without ever having bothered the technician, and this would then be closed out without any further ado, okay? That is one type of example. As we said, omni-channel, Slack is one option. Our next use case we'll look at is device actions. Device actions, as I mentioned, will allow you to do an actual action on the device. Let's take a look together. I was sent a video, and I cannot watch it. What should I do? You can throw in the spelling mistakes and see how it works with it. What should I do, okay? Really sort of classic type of use case. The first thing you'll notice, in order to keep this logged, and a ticket has been submitted inside of a Terra. In the meantime, the AI agent, Steve, as we've called him, is going to go ahead and try to resolve this ticket autonomously, again, without any type of interaction with an actual technician. Now, just like a real technician, as you can see here, it's doing a follow-up. It's going to ask the end user, you know, what's happening, do you have any errors, or what, you know, is going on. Yes. Sorry. Yes, I have an error, and I'm going to go ahead and put in for us today an example of an error that I saved on this video player, and I'm going to send it through to the AI. Okay, again, this is just, you know, this would be mimicking a real use case that each of you might go ahead and meet in your day-to-day. Now, you're seeing here is that the AI is going to examine this. It has correctly understood what the issue is, and it's now going to look for an automated way in order to go ahead and deploy the appropriate video player based on specific applications that have been whitelisted to the AI. It's checking its internal available software. It is now going to the specific device, my device, as an end user, and it's checking to see if the appropriate video player has been installed or not. If it has not been installed, then the AI will be able to go ahead and install that on its own. Now, this works in different fashions. Sometimes this can link you to a knowledge-based article, as in this case, this can just give you the appropriate instructions or whatever the appropriate case here is. As you can see here, VLC is actually already installed, and it's just giving the user the appropriate direction into how to actually, you know, solve their problem. But again, the point here is that in the end of the day, we didn't have to go ahead and actually leverage a technician to go ahead and do this. Okay. So, this shows us a good demonstration of how the end user experience is. Now, unfortunately, my other window I noticed just closed, so I'm going to have to quickly pop into it here. Just bear with me a second, folks, apologize for this. There we go. So, we're now going to show you the back end of how the AI works, okay? You'll notice on the left-hand side, and I think most of you will be familiar with Eterra, you have this AI center, which has been recently released for this. Now, one of the key things the AI can do is to go ahead and examine the tickets that are opened and offer suggestions on how this can be better, or how it, in essence, can automate and remove manual work that you're doing today. You'll notice here, for example, that there have been nine tickets opened in the system on accessing the office guest Wi-Fi network. And it's saying, why are you dealing with this manually? You can use the AI to generate an article, takes all of about five seconds, as you just saw, you can then publish this either for internal use or for your end users, and then you can remove that manual effort, which you're doing today. And that is one type of actionable insight that you're able to achieve with Eterra. Another perhaps almost more interesting one is the following. You'll see here that when, in fact, let's switch over to a different one. As I'd mentioned before, when the autopilot is unable to go ahead and resolve the ticket autonomously, it is able to go ahead and escalate that to a technician, showing the technician, the work that was already done, the diagnostics performed, as well as giving the technician the specific recommended next step. As most of you will be familiar, as a technician, you can then remote into the device, for example, using Splashtop, in order to go ahead and then resolve the tickets. However, if you've gone ahead and remoted in through Splashtop, a specific flag is turned on in the system where the AI is going to analyze your action and see if it can do that automatically next time. Here's a great example. You'll see here where the end user is, excuse me, where the technician has done something for the end user on a ticket that was escalated. In essence, they created a scheduled task using the task scheduler. The AI is saying, you don't need to do that manually. You can go ahead and actually do that automatically with a script. And the AI is able to then take that, with all the information, in about 15 seconds, generate that entire script. You can then publish this for the IT autopilot, as you can see here on the left-hand side of my screen. And the next time this use case occurs, you won't have to deal with this manually, because the AI will do this for you. Another thing I had mentioned was being able to whitelist certain pieces of software to the AI. For those of you who are, most of you are probably familiar, Atera has an integration with three different software catalogs, software repositories. Two are for Windows, that is Chocolaty and Winget. And you can publish applications, as you can see here on my screen, from either Winget or Chocolaty to the AI. Or for Macs, that would be from HomeGroup. So again, this is where you would publish that for the AI. Now there's a little bit more here that we'd like to explore together. You'll notice there's a section of custom instructions. This is where you can train or teach the AI things which are relevant for your organization. For example, as you can see here, in this demo environment, we use a tool called Mesh Payments. But team, this is the training. This is it. The whole kit and caboodle. There's not that much work that needs to be done here, as you can see. You're just telling the AI what you want it to know. Another example I like to call attention to is the following one. You'll see here that Missy Elliott, okay, is a Zoom-connected conference room on Atera's eighth floor that can host up to four people. Now if I, as an end user wrote, I'd like to book some time with Missy Elliott next week For me and three others, if this was a hallucinatory AI, it would probably go on to something like, do you know Missy Elliott? Do you know her manager? Or something along those lines. It would not understand what I actually, as an end user, am asking it. Well, because we've trained it, and again, this is all the training we had to provide, it understands to connect to the 0365 shared calendar and check to see if that room is available. I had also specified it was myself and three others. And if I had said it was a party of 10, it would then perhaps suggest a different room, which would be more appropriate, because as you can see here in its training, it's only up to four people. The last and most powerful one is called Playbooks. Playbooks allows you to customize the default behavior of the AI, the way it reacts, in order to go ahead and work for your specific use cases. But again, the training is exactly that. It's just telling what you want it to do, as you can see here, and then it will queue up the rest of the items on its own. For those of you who may not be familiar, or perhaps are not using part of our ticketing solution, we have recently released the ability to seek approvals through Eterra. This is a playbook that is, as you can see here, leveraging that built-in function, where you would ask the user what's the reason for the request, ask the user who their approver is, it would automatically send the approval request to that person in Eterra, and then assign the ticket back to the technician group to install this application. For example, perhaps they wanted to install Office or Adobe Pro or whatever the case is, something that carries a cost. But team, the idea here is automations, it's efficiency, it's removing the work from your workplace and moving it over to the AI, okay? So I'd also mention other types of integrations. For example, I mentioned about MCP. You can see we already have a catalog of several different options. More are being added as time goes on, as well as other types of integrations. You'll see here some key integrations. For example, maybe you want to have an end user be able to quickly reset their Okta password, or perhaps being able to create a new folder in their Google Drive. Things of this nature are able to be off-boarded to the AI without you having to actually deal with it. So that could be something like my printer's not working, that could be things like I need to reset my domain password, or whatever the case is, you're able to go ahead and off-board that to the AI. Okay. This concludes the hands-on demonstration of the Autopilot. So we're going to head back now to the presentation and continue on with it. And now we're up to the Q&A section. I'm hoping people were putting into the chat. And Alina, if there's anything specific you'd like to let me know, I'll be more than happy to take anything. Yeah, sure. I believe most of the questions have been answered. We still have one question that has not been answered. If you want to take a look from the Q&A section. Sure. Maybe just direct me which question hasn't been answered yet. Since Atera uses system elevation, could end users be able to abuse the AI to make use of higher privileges? Is it avoidable and how? Great question. We in fact had this very recently, where we were able to create a playbook that simply guided the AI saying if the user is trying to abuse the system, for example, they want to get a higher elevated privilege or whatever the case is, simply send an email to the security team and then go ahead and let the end user know that this has been elevated accordingly. Guarantee you they will never do that again. 100% doable with a playbook. Next question is, how does this get recorded as billable time? It's a great question. Remember that every time an end user requests, this would be done as a ticket. In fact, if we go back over to the Atera platform, give me a moment. We can see over here, the tickets. When these interactions are here, you can see all of those interactions. This is the autopilot, as you can see here. We can see that this ticket was dealt with from 224 to 238 in this case, so 14 minutes in this case. While it's not automatically in the time tracking section, you still have that record and can add it in manually. You can also go ahead and filter your tickets, whether it's Steve, John, autopilot, whatever you've called it, to be able to review all the tickets that they have specifically done. That information does get recorded, just not as a billable hour, just now as the answer to your question there. Let's give it a moment or two, if there are any other questions. Feel free to drop your questions while we still have our amazing Uri online to give you all of the answers you need. Okay. Alina, I think I'm going to pass it back over to you. Yeah. All right. Yeah, sure. Go ahead. I was going to say, if there are any other questions that have come up to people's mind as Alina is continuing on with today's webinar, please feel free to continue to drop them in the QA. I'll be keeping an eye out. Now, before we proceed with our survey that I mentioned before, if everyone who's here at our webinar, if you are enjoying Eterra, feel free to share your experience on G2. We would love to hear your thoughts, and let us know how you feel about Eterra, what you're enjoying, and this will just take a few moments, and then you'll also get a thank you Amazon gift card. I'll stay here for a second, just so you can scan the QR code and get into the G2 review page. Great. We have some more questions coming up, if you want to take them. Okay. What is the AI's hardware capacity? I'm asking for the limit of AI so I can understand its limits. So Indra, we can work with basically whatever scale you're after, whether that is dozens of computers, hundreds, or even tens of thousands of endpoints, the AI doesn't actually have an upper limit. It's built for scale. I hope that was able to answer the question. Joshua asked, so back to the potential abuse of the AI, if an end user needs to fix some specific app and the fix requires elevation, will it be handed by the playbook or will it be automated? So the playbook supersedes the built-in behavior. So if they're trying to fix something and you've identified this as a potential area of abuse, you can tell the AI that if they're trying to fix Office, for example, by downloading from the internal repo, other things like Visio, which maybe require an extended license, then you need to go ahead and elevate that to a technician or seek approval and you can't just go ahead and move forward with it. Thank you, Uri. I'm going to go ahead with our survey. So you're going to see the survey about the webinar that you just had on screen. I'm going to launch it now. All right, you should be able to see that. There are obviously no right or wrong questions. However, the system is set up this way that we had to choose one option as the right one. It's not really, so just disregard that message if you see it. We'll be happy to hear your thoughts about the topic that we had today, the rise of autonomous IT. Please let us know your thoughts. All right, thank you so much, everybody, for taking the time today. We appreciate it. I'm going to stop sharing my screen. We're almost – we're three minutes early. Uri, I appreciate you taking the time today to walk us through Autopilot. Thanks, everyone, for joining, and we look forward to connecting in the future. All the best.
