AIT Masterclass: Hands-on Workshop

Webinar transcript

Jason: Welcome, guys. This is not the first slide, my apologies. Welcome to today’s hands-on session, the workshop. For those of you that were here on Monday, welcome back. For those of you who weren’t here on Monday, you should have been, but it’s okay, I’ll forgive you. All of you should have access to the AI. If you don’t have access, please let us know in the chat. I believe all registrants of this workshop should have access to the AI. I’ll show you a few ways to see it when we go into the platform. You’ll also be able to go into the portal and you’ll see some specific toggles within the settings of Atera that we will go through when I get to that part.

# Overview of Today’s Session

Just a bit of what’s coming up today: a series intro, a brief setup for those who weren’t here on Monday. We’re going to go into some simulations, where I’ll be asking certain questions with the AI. I hope you guys will be joining too. I’ll show you how to access the portal, which is how your end users, whether as an MSP or internal IT, will be entering the portal, referred to as the autopilot. Then, we will be in the platform for the co-pilot section. 

# Recap of Monday’s Session 

Jason: Basically, this is a two-part series. The first part was on Monday—a really cool session on setup, making sure email settings are correct, and how certain users are going to get their username and password. That was on Monday. Today is the hands-on workshop with you guys. This is interactive, so if you have any questions, thoughts, or comments, we have Eli here, Meira here, and the full team ready to answer all your questions.

# Introduction to Atera 

Jason: I’m going to briefly go over this due to time constraints. Atera is an all-in-one AI-powered IT platform, divided into our autopilot and co-pilot. We support Windows, Linux, and Mac OS. As you can imagine, we’re an all-in-one platform with AI bolted in, creating almost a new industry in terms of AI-powered IT. If you’re using Atera’s RMM and PSA, which I highly recommend, everyone who is paying for an Atera license has access to the RMM and PSA section. This will hopefully nudge those using a ticketing system outside of Atera to bring it into Atera. This is where the power of Atera comes into play, and adding AI on top of this is the cherry on top. On the left-hand side, we have the RMM, remote monitoring, help desk, network monitoring, BI analytics, and App Center. Bringing everything together, in terms of the AI, we’re using it because we have information about the device, know if they are connected to the internet, on a specific email list, and more. Adding the ticketing system adds a whole new layer of ease of access for end users and technicians. At the bottom, we’re using Microsoft Azure, where our AI model is hosted—enterprise grade, looking at an amazing platform here. Security and compliance: we recently got SOC 2 compliance, which was a huge milestone for us. Data analysis is also available in the form of reporting and analytics tools. 

# Setting Up the Service Portal 

Jason: Let me share my screen again with the platform. I’m going to move to my bigger screen here; it’s easier for me to present that way. Okay, let us begin. First and foremost, I’m going to do a brief introduction into the setup for those of you who weren’t here on Monday. You will see this in the admin section. Once you go to admin, those using the MSP product might see things a little differently. What you are looking for is the service portal. I’m not going to go through email settings today due to time constraints. In the service portal, you will see a whole bunch of settings. For example, here’s the URL that your end users will use to log in. There are other ways using the help desk agent in the system tray that customers or end users can use to log into the service or customer portal. They’ll be directed to this URL. Ticketing is not new. What I want you to make sure you have on is AI assistant. Make sure this is selected within your console. This will activate the AI assistant within the customer or service portal. Do you guys all see this? Do you have this on? Let me know in the chat. 

# Testing the AI Assistant 

Jason: How you can test it is by pasting the URL into your browser. I’ve already done that here, which brings me to the portal. Before entering the portal, it will ask for a username and password. This is not the username and password you use to log into the Atera console. This will be a user within Atera, an end user. If you want to test this, you can create a virtual machine or use another device. For those using the MSP product, you’ll see customers there. Click sites. For me, I’ve used a user named J. This is what I created. You can create a dummy user or add them to a specific customer. Is this clear to everybody? Are you with me? I’ll give you a few moments to do this. 

# Interacting with the AI Assistant 

Jason: Meira, can you pop up the first topic at the bottom of the screen? Thank you. This is where end users will interact with the AI assistant. This is going to be a tier-one technician, taking care of repetitive tasks, like not being able to open a PDF. The AI will check the computer, leveraging the RMM side of Atera, monitoring the agent, seeing what software is installed, and then leveraging that into the ticketing platform. Tickets will show up here once opened. You can set it so that end users will only see the AI assistant and not be able to open a ticket, almost forcing them to speak to the AI first to get initial checks done. This way, when it comes to you, it will only be things that have been escalated. 

# Example: Internet Connection 

Jason: Issue Let’s begin with a simple example: My Internet connection is slow. Remember, this is an AI, so answers might not be the same every time. This helps us think of ways to ask the assistant for help. For this one, we will say, “My Internet is slow.” I’m going to be relatively nice to the AI because we have sentiment attached to tickets, which is very cool. So, we say, “My Internet is slow.”

# AI Assistant Functionality 

Jason: So what’s going to happen is that there will be checks happening in the backend. It’s going to start looking at network adapter information and measuring upload and download speeds. This may take a bit of time because it’s actually performing the tests. These are things happening in the background before you’ve even looked at it. Picture the amount of time it’s saving you as a technician, especially as a tier-one tech. You can see here that it’s looking at network adapter information and measuring upload and download speeds. This may take a bit of time because it’s conducting the tests. Let’s see what the AI says after this. ## AI Response Analysis This is amazing. If you can see here, the AI says, “I see you’re experiencing slow internet. Your upload speed is 27 Mbps and your download speed is 51 Mbps.” I think that’s pretty good. The AI continues, “These speeds should be sufficient for most online activities. However, if you’re still experiencing issues, it could be due to a variety of factors. Could you provide more details about the issue, for example, certain websites loading slowly or trouble streaming videos?” Let’s go and say something like, “When I load YouTube.” 

# Specific Issue Testing 

Jason: This is a basic example, but it shows how we want the AI to work. The AI uses the information we provide to test certain things, so when it comes to you, you can see the tests that have already happened if the issue hasn’t been solved by the AI. The AI tested connectivity to the specified website on Port 80, which is YouTube. It says, “I understand you’re experiencing slow loading times with YouTube. It could be due to a variety of factors.” Then it would provide a solution. Let’s click this and see what happens. Are we sure? Yes. It should have given me a TFA (Two-Factor Authentication). I’ve already done the TFA before this, but it will send a code to a phone. I couldn’t show you this time, my apologies. If it did happen to you guys, that would be cool. Then we can see and load YouTube. Now we can say, “Yes, that worked, thank you.” There we go. You can choose to have this on or off if you want to create tickets automatically or not. Then we can resolve the issue. Amazing. 

# Benefits of AI for Technicians 

Jason: These are all the repetitive tasks, things like turning it off and on again, that you shouldn’t be doing. These are basic things that can be solved, allowing you to focus on more complex tasks. If you’re an MSP, you can look for more customers, or if you’re in internal IT, you can focus on more project-based work. 

# Ticket Creation and Sentiment Analysis 

Jason: Next, I’m going to briefly show you that this would have opened a ticket. We can see that the sentiment was positive. In the next ticket, we’ll try to be more angry and annoyed, to see what response we get. 

# Testing Email Reception Issues 

Jason: Let’s try this: “I’m not receiving emails from [email protected].” Let’s see what it says. It’s probably going to do the same checks. I’m not quite sure if I’m connected to these, but it’s checking listed blocked email addresses, checking if the Office 365 mailbox exists, checking if the domain is blacklisted, and listing trusted email addresses. It’s testing on the backend to see the results of these specific checks. This is where we move away from what you might call a chatbot to a true AI. The AI tests things based on what it has learned to do. 

# User Frustration Handling 

Jason: This could take a bit of time, so let’s see if we have any questions. Actually, never mind, we’ll go right back into it. The AI says, “I understand that you’re having trouble receiving emails from a specific address. Can you please confirm if you’ve checked your spam or junk emails? Additionally, have you recently made any changes to email settings that might affect incoming messages?” Let’s try being angry and say, “I’ve checked everything and nothing works.” The AI responds, “I understand this is frustrating. To help resolve the issue, I would like to run a few checks on your system. This will help us understand if the problem is with your email settings or the specific email address being blocked.” Even if the user isn’t being interactive and is just demanding a solution, the AI is not just repeating questions. It’s conducting tests and checking things based on learned behavior, ensuring a more efficient resolution process. 

# Conclusion 

Jason: This webinar has demonstrated the power of Atera’s AI assistant in handling repetitive tasks, conducting backend checks, and responding to user sentiment. This allows technicians to focus on more complex tasks and projects, improving overall efficiency and customer satisfaction.

# AI Assistant Handling User Frustration 

Jason: Okay, so I agree, James. Very, very cool. Let’s see what we come up with now. I understand that this can be frustrating based on the checks you’ve run; it seems like there might be an issue. I suggest that we try removing the sender’s email address from your blocked senders list. This should allow you to receive emails from [email protected]. Let’s click this, and let’s say I’ve had enough now. I don’t want to speak to this anymore. Okay, let’s put something like that. Let’s see. For this particular case, this is why I wanted to mention that this is the AI. If we see the sentiment here, let’s actually see. We see that the sentiment is negative. This is picking up what’s being said. 

# Sentiment Analysis and Technician Intervention 

Jason: Feel free to reach out. What happens here is that we get the sentiment, which is negative. Now, we as technicians, coming in as Jason, see that the user is upset. Maybe they phoned me and said, “Please help me.” What we can do now is go to what we term the co-pilot. We click the AI assistant on the top right here. This will summarize the tickets. What’s really cool and amazing to see is that you get information helpful for the specific topic at hand. We can then generate a reply from this. Let’s generate a solution and see what we get. 

# Generating Solutions 

Jason: I’m sorry to hear that you’re having issues. This is something you could copy and paste, but we want to try and generate a solution. Let’s generate a solution and see what we get. I see that you understand your situation. This is the solution that it’s already given. We can probably go in and help the end user there. There are times when it will suggest a script for you. Again, this is based on the AI. If it doesn’t work exactly how you would think, remember it’s an AI assistant trying to help you and it’s learning. 

# Handling Repetitive Tasks 

Jason: Let’s try another one. The sentiment we see is negative, which is fine. We can then type a response to that end user, who will see the ticket within the ticketing page. Let’s have another topic. Can’t open PDF. Let’s try this one. Are you guys following me? Are you trying things? If you’re getting cool responses, share them in the chat. I’d be happy to see if you guys are doing things as well. 

# Example: PDF File Issue 

Jason: I realize why it said “trust sender” as a solution over here. That’s because I did the same example before and added them to the trusted sender list, so it was grayed out. Let’s see now. I showed you the sentiment. The sentiment works on the latest reply from the end user. It will change if the end user is frustrated and then calms down. The sentiment will change after each reply. Let’s do, “I can’t open PDF file.” Let’s do this nicely. I can’t open PDF file. Let’s see. 

# AI Assistant for Autopilot and Co-pilot 

Jason: What I’m doing here is what we call the autopilot, where end users interact with the AI. The co-pilot is when it’s coming into the portal and you take over. Are you getting any error messages or can you let me know which software you’re using to open it? I don’t know of any software to use. This is going to be one of the options linked to installing approved software. It’s recommending Acrobat Reader. We can click there and it will install. I’m not going to do it now because I’ve already got it installed, but you get the idea of what’s happening. 

# Complex Issues and Escalation 

Jason: Let’s get to a more complex one. Meira, what’s the next topic? Can’t unzip. We can try that one, but first, let’s try one where it needs to be escalated. Adding to a distribution list. Let’s try that one. 

# Example: Distribution List Addition 

Jason: I want to be added to the distribution list of R&[email protected]. Most basic tasks can be done by the end user. For tasks like this, where the end user might not have access to Azure, group policies, or specific mailboxes, the AI will know and escalate the issue. Let’s see what it does here. The AI should say, “I understand that you want to be added to the distribution list. I don’t have the ability.” It recommends reaching out to your IT department or the owner of the distribution list for assistance. 

# Technician’s Role in Escalation 

Jason: I’m on the IT platform, so I know I’m the IT manager. I can go into the ticket and see “Add to R&D list.” We click the AI assistant and generate a solution from there. It’s given me the same response here, which might be due to using my demo account. If you got anything different, that would be great. Essentially, what you are seeing is that we’re doing tests. If the action can’t be done by the end user, it’s escalated to a technician. 

# Example: Password Reset 

Jason: Let’s try the next topic. The next one was password reset. Hi, I want to reset my password for MS365.

# Sentiment Analysis and Password Reset 

Jason: Cool, cool. Let’s see what it says. Okay, it’s just going to go directly to reset the password. This one is kind of easy. Let’s check the sentiment for this one. It’s neutral; I agree with that. I just said “Hi” and “Reset my password.” Let’s see if I go, “That didn’t work.” What I want to demonstrate now is that the sentiment should change. Cool, let’s see. It should change. Let’s refresh. It’s still at neutral; it should change to negative. It does take a few seconds sometimes, so let’s just wait for that. If it doesn’t change, it’s the AI model that we have, so I don’t really have control over it. 

# Script Generation What 

Jason: I want to do now is demonstrate the ability where it will generate a script for you guys. Let me go to one of the examples I already have. Let’s try a slow internet connection issue. This one was actually closed, my apologies. Let’s try another one, maybe a slow computer issue. Let’s generate a solution. It’s not giving me a script because I don’t have a device linked. Sometimes, you will see that it will give you a “Generate script” option. Let me show you where to find that. If you get an issue that requires a script, the AI will determine if it’s scriptable. It will send you to the admin section. 

# Generative AI Script Generator 

Jason: Go to admin and then to scripts. This is where it will bring you to create a script. If you weren’t aware, we have a generative AI script generator. It will put the topic into the fields and generate a script for you. For example, let’s say “Download and install Foxit PDF reader.” Here we see a script, which is really cool. You do need someone with experience in scripting. Don’t blindly start scripting from the script generator, but the time-saving is amazing. You get the basics out here, and all you might need to do is tweak and define. We also get a script name, which is new. We typed it in here and got a whole bunch of information happening at once. We’ve given it a name, description, and script all at once. It’s really cool. 

# Highlighting AI Features 

Jason: I’ve highlighted most of the areas of the AI in terms of ticketing. We’ve looked at sentiments, different scenarios, and how things can be escalated. We’ve spoken about a few different scenarios. 

# Additional AI Features 

Jason: I want to show you where else we’ve inserted AI to help you guys. One of the ways is as a command prompt generator. Let’s go to Powershell. Here, the idea is similar to the script generator. For example, we can say “Show all open ports.” This is just the basics. These are places where we’re implementing AI as part of the co-pilot side to help you guys and save time. 

# Presentation and Q&A 

Jason: We have 18 minutes left, so let’s jump back into the presentation and then into questions. We have Eli here to help answer a few of these questions. Meira: Jason, we can’t see your takeaway screen. Jason: My bad. One second. While I’m pulling that up, I want everyone to know that we are aware of all your questions. We’ll answer them live afterward so everyone can see the answers. 

# AI Takeaways 

Jason: These are the takeaways: AI is rapidly evolving, and we’re leveraging it in a big way. We’re pretty much the only IT management platform leveraging this revolutionary industry to take our platform to a new level. We’ve invested heavily in developing Atera with AI at the forefront. Wherever we can implement AI, we do, to make your lives easier. Our goal is to supercharge IT with AI and solve repetitive problems, allowing you to focus elsewhere. For businesses, AI adoption is no longer a matter of if but when. Early movers will benefit from an exponential boost in efficiency, accuracy, and productivity. AI is not here to replace anyone; it’s here to open doors for more customers and bigger projects. 

# Questions and Answers 

Jason: Let’s go to the Q&A. Eli, are you ready? 

Question: Can we white label the AI assistant, change its name and signature? 

Eli: Yes, you can. In the customer portal on the top left side, you can have your company name and logo. 

Jason: Let me show you quickly. Go to admin and then to White Label. This is where you can add your logo to the service portal, company icon, and report. 

Question: I keep getting messages like “AI is taking a break” or “Open AI is taking a little break from requests.” Is this a common issue? 

Eli: This is likely due to the configuration of the user linked to the agent. We can help you solve this after the meeting. Schedule a meeting with us, and we’ll help you configure it correctly. 

Jason: You can also open a support ticket, and they will direct it to us. 

Question: Can we get the AI system to install a program, for example, Seven Zip? 

Eli: Every software that is part of the self-service will be available to the AI. The AI will have the option to install it. 

Jason: In the admin and service or customer portal, you’ll see software. You can add software that will be recommended by the AI. 

Question: How much customization can be done to train the AI on guides and checklists? 

Eli: This is something we are planning to do in the future. The AI will take information from the knowledge base and present it as part of the conversation. 

Question: Are the app installs done in silent mode? 

Eli: Yes. 

Question: Are the AI-initiated troubleshooting tools run by the Atera agent on the user workstation? 

Eli: The health checks and applications are based on the operating system scripts, so it should be safe. 

Question: Is there an option for an end user to bail out of having the AI front-end the issue? 

Jason: Yes, you can choose whether you want to have the new ticket option available. Go to admin and then ticketing. You can hide or show the new ticket button. 

Question: Is there a way to see what scope of fixes the AI has the ability to run? 

Eli: Currently, we don’t have this feature, but the AI optimizes the tests and solutions for the user’s problem. 

Question: Can we add or remove items from the AI assistant? 

Jason: Yes, you can add or remove items in the admin and service or customer portal. 

Question: When selecting a new ticket from the desktop tray icon, can we force AI-only ticket generation? 

Jason: We will take that question and email you the answer. Thank you, everyone, for joining today. We hope this was helpful. Meira, please launch the survey. We appreciate your support and hope to see you in future webinars. Thank you! 

All: Bye-bye!