AIT Masterclass Hands On Workshop & Scripting

Webinar transcript

Meira: Great, so again, welcome to today’s AIT masterclass. This is the hands-on workshop plus the scripting. Just to reintroduce myself to anyone who’s new here, my name is Meira. I’m part of the customer success team here at Atera, and I do a lot of these AIT webinars. Today, I’m joined by Sarah and Paz, my two colleagues from the CS team. Like I mentioned, they’ll be answering any of your questions, so feel free to interact with them in the Q&A section. 

What’s on the agenda for today? We have our series intro, our Atera AIT brief, a setup brief, some simulations which include some new features and functionality that we have in the AI, as well as Q&A. Plus, you know, scripting is included in those simulations. 

As some of you will know, and for anyone who’s new and joining the series for the first time, we usually have our AIT masterclass series broken up into two parts. On Tuesday, we had the intro, which is more about explaining where Atera stands with AI, positioning us as AIT leaders. Then we have the hands-on workshop today, where we actually get into the platform and do more of the hands-on demo and functionality. So, that’s today’s session, and again, like I said, scripting is included.

Just taking a look at Atera as an all-in-one AIT-powered platform, what does that look like for you guys and what does that look like for us as a company? Of course, in terms of our AI functionality, we have the Autopilot and the Co-Pilot. Just as a refresher for anyone who’s new, the Autopilot is our end-user facing functionality of the AI, whereas the Co-Pilot is the technician-assisted portion. We’re servicing Windows, Linux, and Mac devices, and we have a whole bunch of different verticals where we operate. Of course, one of the big ones is RMM (Remote Monitoring and Management). We have our PSA side, our help desk and ticketing, network monitoring, BI analytics, and of course, our amazing App Center. We’re hosting our AI on Microsoft Azure. We’re partners with them, and we work very closely to get the earliest features upgrade as soon as possible and really work with their dev team to get the full value out of OpenAI. We’re classified as an enterprise-grade platform and we offer security and compliance. We received our SOC 2 certification, so if anyone needs that, you’re welcome to chat with our support team about that or reach out to your customer success manager. ### Data Analysis and Common Issues We’re just looking at some top five time-consuming requests or issues that we know from speaking to people in the industry, the data that we manage within Atera, and understanding the nature of the work that you guys do. Some of the top five issues that continuously come up are: – 26% reports access and permission issues – 24% email – 21% hardware (printers, scanners, telephones, and others) – 16% software – 13% performance and connectivity 

Okay, so now we’re going to get started in the platform. I’m going to move over to my Atera demo account and we’re going to get started. Before we do, I just want to check if there are any questions that need to be answered or if anyone has anything they want me to discuss before we move on. Meira: I see Sean here has mentioned that you signed up for the webinar earlier this week and you were traveling. Is there a way to get that video? Yes, all of these webinars are recorded and put on our webinar page. I can send you the link for that so you’ll be able to watch it. Jeff: Is it going to apply only to those that have the designer version? Meira: I think you mean the design partner version. So, you all, by being here, should have access to the Autopilot and the Co-Pilot. If you don’t, like I mentioned, just put in the chat your email address that’s associated with Atera so we can add you in. Then you should be able to follow along. Like I said, if you’re not able to do that today, no worries, you’ll be able to get the access afterwards and test it out. 

The first thing we’re going to be looking at is, like I mentioned, some new functionality within our AI, and it’s going to be the Co-Pilot for devices. For those who have been at these webinars before or for those who have tried out the AI, you’ll have noticed that we have our Co-Pilot for tickets where you’re able to get the summarize and solve, generate a solution, generate a resolution, and also generate a response if you need to generate a script from that flow—all of those good things. We now have that functionality for devices. I’m going to stop sharing this page and share my Atera demo account. Okay, I’m just checking that you guys can see it. Yes, okay, great. So again, for anyone who’s brand new and maybe hasn’t used Atera before, just as an overview, this is the main dashboard where you as a technician would come every day. You can see the latest updates regarding your account and where everything stands. 

If I wanted to see, let’s say, the devices Co-Pilot, I’m going to click over here and you’ll see we now have a column for AI which wasn’t there before. I’m going to just take this random desktop that’s available to me on our demo account. Whoops, I wanted to actually click the AI. I’m going to open the Co-Pilot for it. You’ll see here we have some troubleshoot presets that you can already make use of, and these are all really robust and tested and work really nicely. These are just some ideas to help get you started and show you what you can do if maybe you’re like, “I want to use the Co-Pilot for devices, but I’m not sure where to even begin.” This is for you. Just as a reminder, of course, the Co-Pilot is the technician-facing part of our AI, there to really help you guys as technicians. Also, maybe you have new members of your team who are still getting the feel of Atera and figuring everything out; this can help them as well. This one I really like: “What installed printers are on this device?” I’m sure a lot of you who have issues with printers or maybe are managing multiple devices or you have, let’s say, a device that’s connected to multiple printers on different floors—that kind of thing. Also, “What can I do to make this device more secure?” A whole bunch of different stuff you can do here. I’m going to start off with this one: “What’s the CPU usage and top five running processes?” or “Top running processes?” I’m expecting it usually gives me around five, and how can I optimize this one? I mean, you guys can probably vouch for it more than I can, but I know that this is a really important process and a lot of technicians want to know what’s the CPU usage, see what’s taking up the most usage and energy from your device, and how you can best optimize that.

Something I will mention here is the AI does take a little bit of time to generate these answers. For you guys to know, it’s not a bot that will just spit out an automated answer. We actually have the agent on this device running, scanning, doing all the tests and checks and balances, and then providing the AI with answers and updates. So, it does take a little bit of time. That being said, we’re also in the process of moving to OpenAI 4, so it is going to get even faster. But just for you to know, we’re not expecting it to be instantaneous because we want it to take the time to do the checks, to do the balances, and really scan the device properly. Like I mentioned, we will improve that, but just for you to know, this is not unexpected that it does take a little bit of time.

Okay, so in this case, for some reason, the CPU and top five processes gave a timeout. Something took too long here and was stopped, but it’s also giving us another option of how to optimize CPU usage. We can consider the following actions. For me, I would like to see something else, so I’m going to close this for the meantime and try this again, maybe for a different PC. Let’s see if this one will work. Let’s take this: “What can I do to make this device more secure?” as an example. Great, so now we’re back on. Again, like I said, the device is actively being scanned by the agent, which then provides information to the AI. Here we have some software security suggestions. We can see that our software security that we’re using is active and updated, so so far this looks good. Let’s see if there are any OS updates that we have. While this is loading, I would love to see in the chat if this is something you guys would find beneficial. What kind of prompts would you like to ask? Just so I can understand what everyone’s needs are and the feel of the room. Feel free to write that in the chat.

For some reason, this wasn’t completed. Let’s see why and what’s going on here. Okay, great, so we did get an answer. Based on the completed software security check, the device is currently protected by active and updated AV and firewall solutions. Great. However, the OS update times checked out. Okay, so this is something we’re actively working on and refining. This was working perfectly earlier, but you know, classic as it is, it gets a little bit shy when we have a webinar. 

Something else you’ll notice here, which is new and really coming out soon to customers, which I’m excited to share, is that it’s linking the KB articles from Atera. Again, like I mentioned, the Co-Pilot is a technician-assisted mechanism. Once you have an answer from the AI, if there’s anything else that you need to check further, you have the option to see the KB as well. Let me just stop sharing and restart. Just a minute for me over here. So, we’ll see here that if I now go back into the Co-Pilot, where I have the suggested KB article and I click onto it, you’ll see these are articles related to the answers given by the Co-Pilot. For instance, if I’m asking about threshold profiles or something about connecting printers or scanning OIDs, those kinds of KB articles will be linked. It really helps you as a technician by giving you the summary and the short version of the answer from the Co-Pilot, and then any additional information you need from our KBs as well. ### Co-Pilot for Alerts Next up, we have the Co-Pilot for alerts. Same kind of thought process behind it, but now you’re able to recheck alerts, ask the Co-Pilot to give you an update, run the alert again, etc. I’m going to move out of the Co-Pilot for devices and into the Co-Pilot for alerts. Let’s take this one; I think this will probably be a good one for disk usage. We can see here that this is the device it’s attached to and it’s online. Now, we’re going to have the Co-Pilot analyze this specific alert for disk usage.

Great, so we can see here based on the alert triggered for disk usage being significantly higher than the configured threshold, it’s essential to address this to prevent potential performance issues. Here are the recommendations. It gives us the prognosis as well as what we can do or recommended steps to take. If I wanted to actually install CCleaner, I could do that, and that would then be prompted on the actual device itself. That’s where the signal gets sent to. Again, here we have options from the KBs. Another option might be using a script, and you get the shared script library option to review in the KBs, which is great.

Something I will also mention is if you want to be active participants and real design partners for us and help us refine the AI as best as we can, you’ll see here that you have the option for “Good job” or “Not what I needed.” That helps us. We collect that data and process it internally, which helps us figure out if the AI is understanding the question at hand, running the best reports, performing the best checks, and giving the answers it needs. If you want to assist us and continue to make it better, feel free to give it a thumbs up or a thumbs down. Don’t be shy. 

I’m going to take another one. I think I’m going to take the CPU load as my next one. Let’s see, hopefully, we get a better answer than the previous CPU load. As I mentioned, we don’t expect the AI to be instant. We know it’s going to take a little bit of time as it’s actively scanning the specific device. 

In the meantime, I’ll address some questions. I see Jeff wrote, “Are those on this call able to join the design program or is it still closed?” Jeff, if you’re interested, please send me or anyone who’s interested, please send me an email. I’ll write my email in the chat: [email protected]. Currently, it’s closed, but we’re adding anyone who’s interested to a waiting list. We’re going to be working on adding and updating that process and hopefully getting new people on. That being said, I will just make clear for anyone who wants to be a design partner, we need you to be an active participant. If you want to be a design partner, we’re happy to get that started, of course, when the time comes, but we want to get your feedback. If you’re interested, I will write my email in the chat, and you’re welcome to send me an email about becoming a design partner, and we can add you to the waitlist.

Okay, great, let’s see what we get. It says, based on the alert triggered and the check status, it seems we encountered a timeout. Okay, so we need to have a look at this timeout recurring situation over here. I’ll take that with our team, but just for you to see, even still, we do get some information that helps us understand the nature of what’s going on and how we can fix it. Of course, some troubleshooting steps include restarting the machine. If the issue persists, use the retrieved CPU processes from the shared script library. You can see here we also have another KB that’s specific to high CPU usage within Atera. Adjust the CPU load threshold if necessary. These steps should help.

Perfect, so the next thing I want to show you guys, scenario number three, is going to be on creating a ticket—sorry, creating a KB from a ticket. If, for instance, you have solved or resolved a ticket from a customer and you think, “You know what, this is a really good use case for a KB,” where in the future, end users or customers can resolve the issue on their own, let’s create a KB. I know a lot of you don’t have the time to sit and make the documentation, but this should hopefully help. Something else before we continue, I’ll just mention, I know this actually came up in the other session. I’m currently in the internal IT version of the platform. We also obviously have the MSP version. They’re just slightly different. If, for instance, you’re seeing here for me sites and you would have customers, it’s all the same. The functionality is the same. It’s just, of course, geared specifically to the needs of internal IT teams and MSPs. But in terms of the AI functionality that would be available to you, it’s all the same, so don’t worry about that.

The ticket to KB flow would look something like this. It would need to be a resolved or closed ticket. So, I’m going to change this. Great, we already have that view, and there was a specific ticket I liked here that I’m going to look for. I think it was this one. Here you can see that this ticket has been resolved. Also, on the bottom, just a little bit scrolled down over here, we see “Generate article.” Let’s see what we get out of this one. I just want to see if it’s sharing—nope, I need to just reshare that. One moment. Okay, you should be able to see this now. If I click “Generate article,” this is the second tab that’s open. Oh yes, and now you’ll see here immediately this is what pops up. It gives us “How to prevent Slack from launching at startup on Windows.” It’s already taken this issue from—or it’s already, let’s say, translated the issue from this ticket into a KB article. It automatically does the formatting, which of course you can change, but it really gives you a good basis to come in and edit it as you wish. Here you can see it’s really broken it down very simply. If you wanted to create it, it would automatically create the rest of it for you. You can add priority, anything else that’s relevant, what to display in the section, etc. 

I want to try another one, so I’m going to close this for a minute and go here. I’m going to go back to my tickets. This is the one I wanted to show you: an iPhone wipe request. Again, you can use this in any kind of context that you have, but it’s a really good example to show you. You’ll see here there was a conversation with the AI. It was this end user having a chat with the Autopilot, and it was escalated to a ticket. Now we have it in the tickets here that have been closed. I’m going to generate an article. You can see here it loads, and then we have a ticket. So, how to wipe the iPhone. This is a really interesting use case for end users. For those who are internal IT departments and need to set something up like this for your end users, but also for MSPs, the context would just differ slightly depending on what you need, but I’m sure you guys can see the value in this. It really does come out nicely in terms of formats and visuals, and you can customize it as you need. 

I’m going to stop sharing for just a minute. I just want to come back into the webinar and see if there are any questions or any concerns. I’d also love to hear some feedback from you guys. Usually, we have a very interactive group. I know we do.

If you want to share any comments about what I’ve just demonstrated, if that’s interesting to you guys, if it’s something you would use, and maybe any other prompts that you would like to ask, feel free to write that in the chat. I’m expecting some of our recurring participants, our frequent guests, to send something for us and give us some feedback.

In the meantime, I’m going to be moving over to the scripting part of the webinar. Let’s just make sure this is sharing. Great. Paul: Paul has a wonderful comment: “Techies are not the best at doing KBs. This is a game changer.” Meira: Paul, I agree with you. It’s not that they’re not the best, but also that they don’t have time. I know you guys are putting out a lot of fires, and sitting down to write KBs and documentation is a lot. Aaron: Aaron says, “Great features, would definitely be great to have.” Meira: Wonderful. Any other comments are all welcome.

In the meantime, we’re going to talk a little bit about prompt strategy. For those that were on the AIT scripting webinar last month, some of this will be familiar to you. For those that are new, this is going to be exciting and maybe something a little bit new. This is specifically to talk about prompt strategy in general—general guidelines on how to talk to AI, not necessarily native to Atera, but just in general. It’s relevant for those who have never used AI or scripting before, and for those who want to brush up on their knowledge or get better prompts and better results. 

We’ve split it up into three different verticals that we think are interesting and relevant: 

1. Be Clear and Specific: The more clear and specific you are, the better the output will be. The output is really determined by the input, so the more clear and specific you are, the better response you’ll get from the AI. 

2. Length and Context: Don’t be shy about adding as much information as possible. The longer it is, usually the more context the AI has, and the better the output will be. Utilize it to your benefit—give the AI as much information as it can use to give you the best answer. 

3. Time to Think: It’s okay to let the AI think. It’s okay to let it take its time, use all the length and context and specific details you’ve given, and work out an answer that’s best for you. The same would be if we had to ask a person a complex question—we wouldn’t expect an answer within a second. The time saving here is still twofold because it’s going to cut down the time that it would take a human to do the same thing and solve a complex question pretty fast. 

We have two script use cases that we’ve chosen for this webinar today. In the last one, I think there were about five, but today we chose two specific ones that we felt would be useful for you guys on a daily basis. The possibilities are endless for what you can do with our script generator, but we wanted to specifically highlight these two. For those who have tested it before, I’d love to hear your thoughts on it. Maybe people who were in the webinar before have seen them, but these are really interesting use cases that you would be using on a daily basis. They show time saved or complexity broken down really easily and quickly, making you more efficient. 

For smoothness and simplicity, we’ve recorded these flows. I wanted to make sure everything was working as best as possible and be respectful of your time. We show you all the steps that go on for the process both within Atera and outside of Atera, so you get to see the full picture. 

The first one we’re doing is the Chrome installation and update. A common use case here would be maybe there’s a bad version of the app that’s been deployed—it could be Chrome, Slack, or anything. It could be due to an end user downloading a bad version or maybe some auto updates. Some software updates aren’t always amazing, and auto updates may not encourage the best version. Or maybe there’s a corrupted installation. What can we do to address this problem? One of the easiest and most time-efficient ways is to write a script. Not all of us are scripting experts or have the time, patience, or energy to do that. Luckily, we have our script generator.

I’m going to play that for you guys and show you all the steps. As you can see, we go into our script generator, give it a prompt—it’s concise but specific and gives the details of what we need. We create that script and then go to the actual specific device that we need to run it on and run it. We can see that the script actually succeeded without needing to go to PowerShell to run it, but it worked, so that’s great. 

The next one we’re doing is quite a fun one: script proofing, using the specific use case of Slack installation. This is something I recently learned myself. Those who were on the AIT scripting webinar last month will remember we had Eli, our head of IT, join the call. We saw how you could actually proofread a script—take a script made within Atera using our script generator or take a script outside of Atera and refeed it into the AI script generator to check for inaccuracies or fix a script that’s not fully functioning.

I’m going to show you that as well. Again, just to explain the need here—it could be if anyone in the IT team needs to check a script before they’re going to run it. It can work for existing scripts, which is the beauty of it, making it really unlimited in what it can do. We create a script asking for the specific use case to install Slack. We test it out in PowerShell and encounter an error. We copy that error and put it into the script generator. You can drag down and make it longer, then paste the error in here. Our script editor or script generator remodels or fixes the script that’s not working or where there’s an error. It does take a little bit of time because it’s downloading heavier software, but it works. 

don’t know if anyone knew about that before, unless you were on the webinar last month, but feel free to let me know in the chat what you think. People who saw it the first time in last month’s webinar really liked it, as do I, because I think the possibilities are endless. Paul: Paul has a great GIF underneath here that’s wonderful. Anyone else is free to add comments in the chat. 

The use cases are endless because you can take existing scripts made outside Atera and check them. Obviously, the ones made in Atera shouldn’t have any problems, but if there’s maybe a corruption in some part of the download, you can check for it or some kind of error. It also enables your more junior techs who want to get into scripting to proofread and double-check their scripts before they deploy anywhere, which is a best practice and something you should always do.

I want to just cover some takeaways over here and then we’ll get into any outstanding questions. As you all know, AI is rapidly transforming the modern labor and technology landscape. It’s everywhere, and you can’t get away from it. We see it everywhere, and those who get on board with it will just be in a better position. The sooner you adopt it, the better. As an innovation-first company, Atera’s AIT is a decade-long endeavor. Those who were at the webinar on Tuesday will remember that we got our first patents for AI back in 2017. It’s not like AI came into action and we decided to hop on the train; it’s been something we’ve been intentionally investing in and exploring for a long time. We’re in a very good position to be leaders in the AIT industry. Our goal is to supercharge IT with AI’s autonomous and supportive capabilities. We aim to make you more efficient, help you do your work faster and easier, and take care of the lower quality but higher quantity of requests that you get. A lot of FAQs, printer problems, resetting passwords—those kinds of things that you may not want to do. The ticket-to-KB flow that I showed you earlier is really helpful there as well. For businesses, AI adoption is no longer a matter of if but when. Early movers will benefit from an exceptional boost in efficiency, accuracy, and productivity. The sooner you get on board, the better. 

I think we can head over to Q&A. I’m going to see if there are any questions here that haven’t been answered or any that we can revisit. Feel free to share any now in the Q&A section or add any comments in the chat. What have you guys thought about this webinar? Have you enjoyed it? Has it been interesting? It’s been a mix of the usual workshop combined with scripting, so I’m curious how you guys have found it. I’m just looking here to see if there are any unanswered questions. It seems like there are none. Usually, we have a very chatty group. I want to make sure everyone is fully on board and has any questions, concerns, or thoughts. I’ll give it another minute or so for you guys to add anything else. Sam: Are the webinars posted anywhere? Where could we see previous ones that we’ve missed? 

Great question. I’m actually going to send that to you now. We have all of our webinars saved and uploaded on demand. They are all recorded, saved, and then uploaded so you can watch them. I’m adding this link in the chat right now so that you guys are able to follow along. That should work for you. Karen and Chris,  Thanks for your comments. Very informative. Happy to hear that and happy to have you joining us. Brian, Are some of these features we saw today with the Co-Pilot close to being available to all? I believe so. I need to double-check on the timeframe, but I believe they are in the process of being released. I need to check if it’s first to design partners and then to everyone. But it’s close enough that I can show it to you guys and get you excited. Really good question, Brian. Karen, If you don’t mind sending, I’m just going to put my email again here in the chat for anyone who wants to be added to the design partner program. Feel free to send me an email. Just make sure you include that it’s for design partners. We are happy to add people to the waitlist and eventually get you on the design partner program. But we want customers who are going to give us feedback because that really helps us adapt, refine, and tune the product to make sure it’s as best as it can be. Sean, It’s my first webinar and it’s been easy to follow along and good info.

Well, Sean, big welcome to your Atera webinar journey. We hope to see you again. We hope to have you coming back. We do have a lot of recurring attendees for these webinars and a great community around it, so I hope you’ll be back with us next time. Aaron, Thank you for your great comment. Aaron is also one of our recurring guests, so it’s always lovely to have you with us. Any other questions? It seems like not, so I’m happy to give you guys some time back. I’m sure all of you are very busy and have taken out some time of your day to join us, which we’re so happy to have you. 

Before we actually sign off, one last thing. If you don’t mind, I’m going to launch our survey and get your feedback. This really helps us refine these webinars, make sure they’re interesting and relevant to you. The scripting webinar was highly requested from feedback from all of you who join us. The more insight and information we get from you, the better we can tailor these webinars to what you’re interested in hearing about and looking to see. I would love to hear from you guys. It shouldn’t take more than a minute, and we’ll all be on this session together, so you’re not alone doing it. While we’re doing that, I will just close up in the meantime and thank you all for joining. Like I said, I know you have very busy schedules. We have such a wide variety of attendees today, so whether you’re joining in the morning, in the middle of the day, or staying up in your evening to join, we really appreciate it. We love having this very supportive community of our AI webinar attendees. I would also like to thank Paz and Sarah, who did an amazing job in the Q&A and answering all of your questions. I’m happy to give you guys back your time. Feel free to finish up the survey, and we will see you at the next webinar. 

Bye, everybody!