Community town hall – February ’25

Webinar transcript

Gil: Hi everyone, uh Gil, actual name Gil Gon. I go by that name because, as probably most of you know, there’s another Gil in charge, Gil Pekelman, a different guy. He’s also really great. This is our second Community Town Hall. I’m very happy this is becoming somewhat of a recurring thing. We’ll have a ski for the next quarter since we have a bit of a bigger event. It’s going to be the Terel, and there’s going to be a session that I will be involved in directly. So that will be kind of our next Meetup.

But for today, we have someone pretty special here at Thea, our CTO, co-founder, and a very good soccer player. He’ll tell you a bit more about himself. I’ll just share my screen. ## Housekeeping 

Gil: Awesome. So yeah, today we have an opportunity to introduce you to our CTO, and he’s going to talk a bit about our AI Co-pilot Innovations and show you some of the very slick things that he’s been working on with his team. Let’s start off with some housekeeping rules. This is a recorded session. I know that everybody likes to ask, and I’ll repeat it if someone asks during the session. We will be sharing it later on in the community, and you’ll also be able to see it on the VOD area in the webinar’s page. If you have questions, please, by all means, we love it. Just make sure that you’re placing it in the Q&A tab and not the ongoing chat. 

If I see a question there that pops up, I’ll try to remind you of that. Just be aware, and at the end of the webinar, there will be a short survey. It’ll be launched, it’ll take about a minute to fill in, and we’d love to hear your response. It really helps us as we move forward catering to the things that you want to know and see more about. And a reminder to sign up and sign in to enjoy the full perks of the community. Make sure that you’re not just scrolling around; you’re actually signed in. That gives a better indication, and it means that whatever you do, you’re able to win your perks and rewards for it. There’s a quick QR code here for anybody that wants to check out the community as we discuss or for later on. ## Overview of Atera’s Platform 

Gil: So, kind of when we talk about Atera, just so that we’re not missing out on anything, it’s an all-in-one platform, and really we offer an extensive suite of solutions for IT professionals such as yourselves. We talk both about the platform, which includes RMM and Patch Management and Remote Access. There is the PSA area such as the Service Desk, the Ticketing. We have an extensive App Center where you can add different integrations, and hopefully, you’re using all of these different features. All of them get updates all the time. I try to share as many feature updates as possible, but even if there’s an area that you haven’t tried for a while, I really recommend heading over and seeing what’s new and how we continuously develop it. 

At the heart of all this, as you can see in this little image, there’s now Autonomous AI, which is the powering force behind our AI. Today, we want to dive a bit more into the Co-pilot area of it. So, I’m talking about Autonomous AI. There are multiple instances where it integrates directly with the platform itself. The same areas that I was talking about earlier that are relevant for your day-to-day work are the areas in which AI Co-pilot works. It’s an add-on; you can access it through the marketplace, you can activate it directly through an app. The idea that we wanted to offer today is to see how the different areas can work for you and how the autonomous work is the future that is going to make your life better using AI Co-pilot. 

Of course, I always like to encourage use, and I like it when you’re able to show off. So we have some AI Co-pilot badges, and for any one of these different areas, I want you to mess around, try it out, see if it works for you. Give me a ping, and you’ll be rewarded with the badge. Earn the badges, get the swag. I know that a lot of people think it’s pretty neat. Of course, this is not the main reason to use AI Co-pilot, but I always like to add a little something. ## Introducing the CTO 

Gil: And yeah, with that, the main event for today. Let’s talk a bit with Oshri here with us. So, Oshri, other than playing football, what else can you tell us about yourself? 

Oshri: First of all, thank you. Thank you for hosting me here. I think it’s my first time in the community, so I’m very excited. 

Gil: Thank you. It’s a pleasure, and thank you so much for taking from your time to be with us. I think that the perk of being able to ask questions to one of our leaders is something that is really nice for the community and doubles down on the belief that you have in the community. So thank you so much for having this area and the product. It’s not something that all SaaS software has. I appreciate it a lot. It gives me a place to work and a place to have fun, and it gives people an opportunity to connect with you directly. So, thanks for attending today. ## Demonstration of AI Co-pilot Features 

Oshri: You’re welcome. So, I think you have a lot to show. If it’s fine, I’ll stop sharing, and then you can take over.

Oshri: Great. So, actually, before I start sharing my screen, we’re talking about Co-pilot and all the AI capabilities. I’m going to show you things you might have seen in the system, but I want to give you my point of view and some pro tips, and then we’ll move on to the new stuff, things that we are about to release or are in the development process. So stay tuned. We’ll start small with some things you might know, but as long as we progress with this session, you’ll see amazing and very exciting things that are coming next. So I’ll share my screen. Just a sec. 

Oshri: Yes. Okay, we’ll start with the device list. I’m not sure if you all noticed, but we have here the “Ask AI” little button. It’s a bit confusing because the default is to search. So if I want to search my computer, I can just click like this. But once I click on the “Ask AI,” you can see a little bit of an explanation here of what you can do. You can ask very nice questions. You know, it’s actually doing a lot of work for you behind the scenes, filtering this device list, and then you will get a list of devices you can act and do actions on in a multiple way. I want to show you one very nice example I have. Let’s ask something like “online unpatched Mac machines with 8 GB RAM or more” and just click on this button. You can see we get the list immediately. I can, of course, click the “select all,” run script, assign automation, do something. But look at this little sign here. You can see that it’s modifying my filters. 

Of course, it’s Mac devices as we wished, online devices also something we mentioned in the query. But we have advanced filters that most of our users actually don’t know even exist. If we look at this, we can see that it filters devices with memory equal to or greater than 8 gigabytes and machines that are not fully patched. We have a lot of different filters we can do, but you know why doing it manually one by one? Just let the AI do the work for you. Just ask whatever you want. I’m going to say that behind the scenes, every time you guys look for something that we don’t have, you will get a notification, but we will also get a notification. Behind the scenes, we are collecting this feedback and improving our system. So feel free to ask whatever you want here and stay tuned because we are doing a lot of work behind the scenes to extend it all the time, and we’re watching exactly how many queries we weren’t able to give you the results for, and we are adding more and more filters to this system so you’ll be able to use it and enjoy it. 

Oshri: The next thing, let’s move on to the script area. You can write in the chat if you’re using a script generator and scripting in Atera. It’s interesting to know. I want to show you something. I’m not sure everyone knows about it, but when you create a script, you have the script generator. This is not new. I think we have it for one and a half or even two years now. You know, we started early doing AI and implementing AI in our system. So I want to show you something more of a pro tip in this area as well. There is a journey I want to take you on with this inside Co-pilot in Atera. 

Oshri: First of all, what I want to do is create a script that lists the top 10 largest files on the desktop. You can see it’s immediately generating a script, but I want to show you something. This will gather all the files from my desktop, select the top 10, and sort it by size. Then we’ll do something. Actually, I want this script to be a cleanup script, but for the sake of this demo, I won’t delete any files. I’ll just list them, but we’ll call this script “Desktop Cleanup.” Okay, and I want to show you something before we move on with this script to Co-pilot. I can just go inside the script, let’s say in the middle of the script, and I can add a comment, something like this: “Ignore MP4 files.” So, I want to list all the files on the desktop or delete some files, some large files on the desktop, but I don’t want to delete MP4 files. So if I write just a comment and then click enter, you’ll see immediately the AI will work for me and suggest the next lines of code. So, as you can see, it filters out extensions that contain MP4. Okay, and I can add more and more comments. Every time you click a comment, you just tell the AI, “I want you to do something in the next line.” 

Oshri: Don’t list files with “do not delete.” Okay, now the AI is thinking, as you can see here in the small icon, and it’s suggesting new lines of code. As you can see here, I’ll just click on the tab, and it will accept the change, and you can move on. So it’s not only a script generator, but you can do inline editing, and the AI can help you modify the script and add some more logic into it, also into existing scripts, by the way. So let’s take this script, and I want to show you something very nice. We call it “Desktop Cleanup from Large Files.” I’ll just click on this button. I don’t know if you’re familiar with this one, but enable for AI Assistant. We have Autopilot and Co-pilot. For our demo, Co-pilot will be enough. So from now on, Co-pilot is now aware of this script. Because it’s a desktop, I need it to run it in the current user, not as a system, because this is a variable environment variable that only users will have. Now we’ll save this script. We called it “Desktop Cleanup.” Let’s give it a better name: “Disk Cleanup from Large Files” and click “Create.” 

Oshri: Now, when we clicked on the create, you can see that in the script list, you have “Allowed for Co-pilot.” So this is exactly what we did in our script. We are allowing Co-pilot to read the scripts and use them. So now, if I’ll go back to my device list, I will look for my machine. Just a sec, just remove the filters. This is my machine. I’ll just click on Co-pilot and say, “You know what? Can you help me clean large files from my desktop?” Now, Co-pilot is actually aware of the script, and he should offer me the script. Usually, it’s also taking some time for the system to index the scripts, but it will be available very easy. You can see I found a solution to help you clean large files, and you see the script “Desktop Cleanup from Large Files.” I just click on it, and that’s it. So you guys can create whatever script you want. I saw in our system that the IT guys did some changing screen savers, cleaning up scripts, patching scripts, and as many scripts as you can imagine. Then you can just ask Co-pilot to do it, or even if you ask Co-pilot to fix some problem without giving him the solution, if he identifies that your scripts that exist in your repository can solve the problem, it will offer it like I just did. I just said, “Help me clean large files from my desktop.” He’s identifying that I have this script in my repository, and just with one click, it will do the work for me just like that. ## Q&A Session 

Gil: Cool. So if we are in Co-pilot right now, guys, if you have any questions, feel free. We can do it dynamically like this before we move on to different sections. But I just wanted to say that if we’re already in Co-pilot, once we go into a new chat, we are not attached to any device. We can just click on the device dash, and you can select any device that you want like this and start working on it. But what I wanted to show you here is that you can just ask some generic questions like, “How many Mac devices do I have?” Something like that. It’s actually connected to the device data. It’s also connected to the ticketing data. I must say that today we started with this very small subset of data that we gave Co-pilot access to, but we are going to do much more than that in the future. So today, we can ask about devices, customers, and tickets and things like that. I’ll give you one more example. I can ask something like, “Find two open tickets that you believe a junior technician can easily solve.” 

Just like that, Co-pilot can query our ticketing system, find the tickets, analyze them, and give me the answer to this question. So we have a docking station malfunction issue and a guest Wi-Fi password. Yes, probably those are very easy to solve. The Wi-Fi password should be an easy one. Probably the docking station is not connected or something like that. You know, it’s also reading the content, so you know exactly why he gave me those results. So feel free to use it. As I mentioned about the filters in the device list, if you don’t get something, we get the feedback behind the scenes. So we’re connecting it to more and more data endpoints behind the scenes and will make it even better over time. So start using it. It’s not perfect yet, but it will be great. 

Gil: Actually, we have a question that I think that maybe now is a good time for it. Sorry to interrupt. Joe is asking, “I like what I see. It’s made me a better tech. How do I start?” So you’re giving these good examples, different cases. What would be a starting point that you think is relevant using this? Oshri: Actually, I have a very nice demo in two minutes. We will get there, that will show you that you don’t need to do anything. The AI will help you start. We are going to introduce some new areas. You can see it here, AI Center, that before you start, the AI will watch everything you do, how you answer tickets, and then suggest what would be the next step to do. So stay tuned. We’ll get there in a sec. 

Gil: There’s another question. Would it be possible to run a script via autopilot? I’m guessing that maybe you’re referring to the Co-pilot feature that we saw previously but on a different server. I think there’s even an example here. A user is having its own computer and working also on a terminal server. Something in his terminal server session is going unexpected. Could he run that via autopilot or Co-pilot on a different computer? 

Oshri: So it’s actually related to something that we are working on, that a user will have multiple devices. So this is in the works right now, and we are on it. We do have a new feature in autopilot that when a user asks for something that exists in the script, you can mention in the script that it should run on a different machine. Usually, this case is most common when you need to run something on the local domain. So we are adding it to autopilot right now, but I think it’s referring to the first one I mentioned, that a user needs to have multiple devices. 

Gil: Awesome. Thank you. And Matias, if you want to expand on that, you’re most welcome to in chat. ## Custom Fields and Automation 

Oshri: Great. I want to take you to a different area in the system right now. You’re probably familiar with custom fields. We have in the ticketing area, you can go to a ticket and create some custom fields. I want to show you an example I did two minutes ago, but if you have any idea of creating another field that you guys are interested in, just let me know, and we can do it live. I created this field called “Onsite Visit Needed.” Just a small checkbox field that I want to have in my tickets every time one of my technicians needs to go on-site or visit someone on their machine, and this work cannot be done remotely. Now, we created something very nice in automation, in ticket automation rules. Gil, this one is new, is in development right now, so it’s very exciting. We created this “Set Field Value by Co-pilot.” 

So every time a ticket is created, or the technician, you can actually do whatever triggers you decide. Every time a technician comments or a customer replies to your tickets, or status changes, you can say under these conditions, “I want to do this action,” and the action is “Set Field by Co-pilot.” As I mentioned, like this, “Set Field by Co-pilot,” and you can select whatever fields you have in the ticket and also the custom fields that you just created, like I did, “Onsite Visit Needed.” From this moment on, I did this trigger every time a new ticket is created in my system, go and check the ticket content and decide for me if an on-site visit is needed. 

Oshri: Now, I created two tickets in my list here. I created one just right now, “My computer is slow” by Ion. If we just go into this ticket, you’ll see that the box is not checked. And I created also this one, “I need a new display, mine is broken.” Let’s log into this ticket, and you can see that the AI set the “Onsite Visit Needed.” But this is only my imagination two minutes before this session started. So you can create whatever fields you feel like and take it wherever you guys want and let the AI just fill those fields for you. 

Gil: This is amazing for managing the days, the travel, making sure that you’re utilizing your time right when you’re on-site, especially for those of you that are

Oshri: Absolutely, and I saw that some customers that I talked to about this feature, they were saying—some of them were MSPs running an MSP business—and they were saying, “Let the AI tell me if I have an upsell opportunity here to sell something,” like selling a display or maybe new hardware or backup. But it can also identify where a technician gave a bad answer to your clients or whatever. You can take it wherever, and I want to show you something very nice. We’ve added also into automation something very cool as well that will complete this flexibility. We have the auto-tagging system. So every time you have some condition, some event like a new ticket created, our system is going behind the scenes and tagging this ticket based on the content. But until today, you only had to choose from a closed list that Atera created generally, and you couldn’t change it to whatever you wanted to tag. So it’s in the works now, and I don’t have yet the ability to show it, but we are going to release it in two weeks—a management system that you guys can just remove and add your own tagging and tell the AI, “You know what? Every time a ticket is created, these are the tags I want you to put on this ticket if it’s actually relevant.” If you think it’s relevant, let the AI do the work. It can be, “Do I need to escalate it to someone else?” or things like that. Or, “Does this ticket need some special knowledge?” I don’t know if you have some very unique systems that only one technician can know how to work with, and then you can put it as a condition here and then maybe assign this ticket to a certain group or to a certain technician. 

Gil: This is a great customization option, and it can create escalation rules as well. 

Oshri: Exactly, escalation rules. You can forward the ticket based on that. You can set some fields based on that. So those two give the AI the ability to tag your tickets, but you select what tags you want it to select from, and then let the AI just fill all the ticket fields for you. This is powerful, and I saw a lot of very amazing scenarios where this takes you as a company to another level. Cool, any questions about it, guys? 

Gil: I have a couple of questions here, new ones in chat. In ticketing, a user submits a printer not printing. Can AI run a script to reset the print spooler automatically? This is from Peter Art. 

Oshri: Absolutely, but this part is autopilot and not co-pilot. It’s in beta right now. Today, we have a few design partners that are working with us on autopilot, and their users are getting answers from AI technicians. It’s solving today around 20% of the tickets autonomously and also escalating tickets to the second tier with a lot of information. We have something like three or four hundred actions and health checks that the AI can run, and it’s getting much better from day to day. It’s going to be live in one to two months from now, so stay tuned on that. 

Gil: Yeah, that’s I think another step on the way, and hopefully in the end, we’ll have some time to discuss that as well. Another question from Matas, thank you. Are there plans that AI becomes a condition instead of an action? For example, priority is high but no feedback for three days, so ask for an update or decrease priority. Ticket is not updated for seven days, automatically close it. Steps have been provided for the customer to execute but no feedback, so provide a new comment with more detailed instructions. Customer makes an angry comment, so raise urgency. These are a bunch of different ways that the customer may be responding or not, and how does the AI come into play? 

Oshri: So I think this can be done with our current automation system. As I mentioned, when you will be able to customize the tagging, you can get exactly to this solution as well. 

Gil: Awesome. And yeah, we can continue. Thanks. ## AI Center 

Oshri: Great. So I mentioned that we have something great going on, and we had this question about how to get started. I want to show you something very exciting. I’m going to go back to our environment. Sorry, just a sec. Great. So in here, we have the AI Center. I’m not sure you all noticed, but maybe it’s only in beta right now. But we have a new area called AI Center. First, I might say that before we go into the new stuff, you have the Optimize AI section. Please, please go in it, read the recommendations. It’s asking you to add the knowledge base so the AI will be much smarter. It asks you to assign agents to contact to connect to your Office 365 if you have set the domain with the right credentials so the AI will be able to connect to your domain and run actions on it, like resetting passwords, checking if a user is part of a group, or things like that, and give you more and more very easy tools to do the things that you do on a daily basis. So make sure you have a good number here and go and optimize your AI because this way you’ll get the most out of Co-pilot and Autopilot later on. 

Now I want to show you this tab called Actionable Insights. It will be available in the next few days. And I want to say that behind the scenes, every time you guys are working on a ticket, we have this concept called AI agents. I don’t know if you’re all familiar with AI agents. AI agents are actually AI pieces of software that are running 24/7, reading some of your tickets and comments. Of course, not learning out of it, but it’s just reading it and seeing how you guys in your environment can do much more if you just pay attention to some things. I’ll give you some examples because it’s very hard for me to explain it like this. You can see that the AI identified out of four tickets that I had in the last period that my technicians are replying to our end users on how to troubleshoot Cisco VPN client. 

So they are saying enough is enough. You don’t need to do it all the time and reply with the same reply. Just click on the review article. I’ll create a script out of your answers because the AI is identifying that I four times gave the same answers to my end users in tickets. And now I have a knowledge base article ready for it. So it’s very easy when a ticket is open to just send this article to my client, or if they have access to the knowledge base system, they can just log in. If they are using Autopilot, Autopilot will know how to answer out of this knowledge base article. So it’s also a way to teach the AI how to solve things. 

Gil: This is incredible. It also relieves the time that I need to think, “Would it be quicker to answer it, or would it be quicker to create a knowledge base article?” With AI taking that kind of thinking off, it means that you can do the more important work. 

Oshri: This is why I said before that you don’t need to do anything to start. Just go inside the AI Center. It will just identify everything you need in order to have a very nice system with efficient controls. And I want to show you something very nice as well. It’s not only the knowledge base articles that the AI can help me with, but it’s also giving me scripts. It identified that I did something manually in my system a few times, and the AI is saying, “You know what? Don’t do it manually anymore. I have a script for you.” I want to show you it here. You can see that in the Actionable Insights, I have the list of the articles that it’s suggesting to me, but it’s also suggesting scripts. 

So in this case, and it’s in our internal IT environment, I just saw this insight and I was wondering why we have it here. I went on a few tickets and noticed that some of the devices are usually connected to docking stations, but when they go home and are working with their Wi-Fi, I don’t know why, but the Wi-Fi was disabled. So they called Avi, opened a ticket, and Avi was saying, “You know what? Next time you’re connected to your docking station, I will just enable the Wi-Fi adapter.” And the AI identified that Avi did it a few times, as you can see, six times, and now it’s saying, “You know what? Don’t do it manually anymore. Just click on the review script. I’ll create a script for you that enables this Wi-Fi adapter, and that’s it.” From now on, if you click create, you have this script. 

The AI is giving you this very nice indication that you did a great job in adding this script to the system. So again, if you have Autopilot, this script will automatically run every time a user complains about this problem. But also in Co-pilot, you can say, “Enable this device,” and the next time the user will connect to the docking station, this script will run and solve the problem. 

Gil: This is amazing because, as I mentioned, you don’t need to do anything to start. The AI will just read and see everything you do and keep you up to date, you know, and give you some tips and scripts and knowledge base articles so you’ll be able to do a much more efficient job. 

Oshri: Yeah, this is what we’re talking about when we talk about Co-pilot as a colleague, as someone that’s there with you and not only notices when things need picking up but can also resolve it. So he’s working around trying to make life better for you. Having that kind of efficient colleague that is on it 24/7 and always looking at what can be done to improve your work cycle is exactly the promise of Co-pilot. This is really interesting, and the AI Center is a very interesting step forward. 

Oshri: Absolutely, and this is just the beginning. We started with the AI agents, I think, two years ago, and now we’re seeing amazing results, but we just started. We are going to add more and more features into it and more capabilities, and Autopilot is about to go live soon. So stay tuned, guys. It’s exciting times to be in this area. I am excited. I hope you guys are too. 

Gil: Thank you so much. It’s inspiring that you still see the fun and the things that come up, and then you come into work with a smile every day and are checking what’s new and what we can improve. I love that. Thank you so much. Since we have a few more minutes, firstly, if you guys have more questions, now is a good time for additional ones. ## Closing and Upcoming Webinars 

Gil: I’ll just share a bit more. There we go. You’ve seen AI Co-pilot. Perhaps last time that you tested it, it didn’t have all these features, and now you want to dip your hands into it again. If you want a trial, you can set up in-app. If you’re having any issues with that, just reach out. I’ll gladly help. I’ll also post it again in this webinar post that I’ll upload next week, where you get the full link and the information on how you can try it out for yourself. I also wanted to use this opportunity. We have another interesting webinar coming up next week with Gil Peckman, the other Gil, as I sometimes call him, with one of our customers. 

They’re going to talk a bit more leadership-level, deep-level into autonomous IT, the evolution, and the benefits. They’re going to talk about what it means for IT leaders and the whole industry as a whole, different perspectives in AI adoption, and of course, the different challenges and opportunities. We’d love to see you there. You’re very welcome, and in the webinar’s page, you’ll have the link for it next week. I’m just going to check again for additional questions. 

Gil: Okay, so this is more on a thought leadership level, and I think you have the right person for it. So Matias is asking, “Do you feel that AI has a high priority in Atera? Taking all development into account, how much percent of your daily development work is currently focused on AI compared to non-AI features?” 

Oshri: So, that’s a tricky question because AI is also part of everything we do in Atera. Also, we are using a lot of AI tools to make us much more efficient in our work. But I think that in the nature of our system, we have much more RMM, PSA work other than AI features. But it’s going to be very hard to distinguish between features that don’t have AI and those that do. My guess is that 30% of what we do is AI today. 

Gil: In our previous Community Town Hall, for those that couldn’t attend, we met three of the different PMs. So I don’t want to get too deep into it, but the way that the work structure here at Atera operates is that we have different domains, and the product managers are each experts in their domain. They’re creating these, let’s call them platform features, and thinking about how we can offer you more in a way that on one hand is completely autonomous on its own, and on the other hand, like Oshri said, we’re always thinking about how this can connect with the ideas that AI offers and opens up. I think that’s a general way of how we’re looking at things. 

Gil: Another question from James: Is the Co-pilot you refer to anything to do with Microsoft Co-pilot, or is it your own AI implementation? Oshri, I think this is a good one for you. 

Oshri: Yes, it’s our own. It’s not related to Microsoft Co-pilot. 

Gil: Sweet and to the point. And Brian, I see that you asked a question. Will Autopilot be included with Co-pilot as far as pricing goes? For that one, I don’t mind taking that, Oshri. At the moment, I don’t have the details about this. When we’ll be launching it, and again, wink wink, it’s coming up, we’ll have much more to share about that—how it’s going to work, what it means for people with Co-pilot, what about people that want to join the party. So a bit more patience, and then we’ll have all the information that you might need for that. I’m just going to check if there’s additional questions. 

Gil: James, I see also from earlier on, but I think that we covered now the implementation of it. So again, guys, thank you so much for joining us. This will be available as a resource for later on. I hope that it gives another perspective about how things are developing with AI Co-pilot, what it can offer you today, and what’s coming up tomorrow. Oshri, thanks again for being open and showing us the things that are coming up. And thank you, everybody, for joining us. 

Oshri: Thank you very much.