Transcript
As we go through this, there is forward-looking content that we are going to talk about. So, this is the legal disclaimer. Nothing in here is guaranteed, but what we talk about today is either live or on a near-term roadmap. So, we're excited to share that with you here. And the other legal disclaimer, everything here, all logos, those are exclusive property of Enable. I am the host today. I'm Nicole Renike. I'm the Chief AI Officer for Enable, and I've been part of the team to bring this to life. Our co-host today is Michael Weaver, who is the Product Management Lead in charge of Enzo and all of the other exciting things that you're going to learn about today that are coming out alongside him. If you're new to Enable, you may not know this, but Enable has been an AI provider for quite a while. Some of the things that you may already see in your products are things like streamlined workflows and threat detection and threat remediation. So, all of those items up here in white, these already exist. They're already in the market. And what we're here to talk about today is this little guy over here in green, and that is your in-product personal AI assistant. That's called Enzo, and it's all part of a much bigger vision. Where Enable is going is we are going to be providing AI for all of your automation needs. So, if you think about what you currently do in the Enable products, you have technicians that log into software typically, and they make decisions and they use it to perform activities or to create their own automations. And where we're going as an organization across all of Enable is that we are going to be the systems that collect all of the signals from all of your environments. And a signal is data that's activated and correlated with other information. And we're going to automatically do things on behalf of the technicians, which you have given us approval and human-in-the-loop approval to do. So, things like fixing vulnerabilities and assessing risks, all of those activities that take you so much time today, those are things that we are going to be bringing in the loop and have already brought into the loop to allow your technicians to get on with their day more quickly. So, we're very excited about this. And to make this come to life, we've actually created an underlying architecture that allows us to scale both across all kinds of use cases and across all of our product lines and enable you to expand across all of your vendors. And the way that we do this is with a basic agentic architecture. So, at the very, very bottom, we are able to take in all kinds of data, including your endpoint device information, the intent signals, so the information we gather through like Lumen, your standard operating procedures, and all of our knowledge-based information. All of that data, those are signals that go into agents. And we have AI agents that have expertise. So, they have well-understood program manuals or operating manuals that are associated with MCP servers. And those MCP servers, they have tools that you can take action with, which is really incredible because now that we have agents that are experts and we have MCP servers that are tools, not only can you use the things within the enable architecture, within the enable product lines, but you can use it with your own tooling. So, if you have an agentic architecture with Cloud and you want to access our MCP servers, because of the way we've architected this, you can do that. It's incredible and it works across vendors. So, that's very exciting. So, what we're going to show you today and what is live and has been in preview for over a quarter already are the experts that we call an asset expert and a product expert. We'll talk a little bit more about those in a second, but it's really important to understand that these are experts that have well-understood operating manuals that access your live information. So, it will actually go in, it will understand what's in your infrastructure, understands the permissions of the logged in user, and then incorporates that information into suggestions and outcomes. The way that you see this at first is called Enso, and yes, Enso Ferrari is involved. We are super speedy, it is super exciting, and you can't get away from having that end dash in all of our naming conventions. So, we're excited to give you Enso. Enso is the front end on that architecture we just talked about. So, what you see here looks like a typical chat interface, like what you would see with Cloud or chat GPT, but it's actually the way into that agent orchestrator. And then the agent orchestrator communicates down to each of the experts, depending on the kind of question that you're asking. And what we're seeing is you get smarter decisions come out of it, it's reducing the risk of not understanding something or not finding something, and it really is boosting productivity within the products themselves. Some of the key use cases that we're seeing people find value is understanding, hey, is there a report that already exists to do something for me? Is there a report that I can use? And instead of having to go out to the knowledge base or to dig through things yourself, Enso is trained on all of that information. You can find new ways of saving money. So, where can resources be freed? Which servers are underutilized? Information that's hard to get from a dashboard, we can actually infer the answer from the information that we have access to and give you the answer to questions that you can't just get by doing a filter on a table. Additionally, you can start to understand your customer environments and ask questions about your customers, and you can do diagnostic questions. So, if there's an asset, you get a call about an asset, and it's not running properly, you can say, hey, why is this specific asset not running properly? And it will give you all the diagnostic information and troubleshooting steps, all based on our experience and knowledge base articles. But nobody really wants to hear me talk about what it can do. What you want to see is the star of the show, and that's Mike Weaver. He is going to give you a demo of everything Enso can do and all of the things that you can do with direct access to our MCP servers. So, Mike, take it away. Great. Thank you. It's exciting to be here, and I don't know if I'm the star of the show, but I guess I get to show the great work here that our teams have done. So, let me share my screen, and we'll start with Enso. Great. So, I'm going to show you an inside environment where we have Enso set up. And so, right away, right in the top here, when we start the preview, you'll see Ask Enso here, and you'll get our lovely assistant here, and with some suggested prompts. A little trick when you're doing the more complicated items, you want to go ahead and make it nice and big so you can see it full screen. And this is a great assistant. This allows you to ask in plain language what you're looking for, whether it's documentation, anything about the assets that you're managing. And as time goes on, as Nicola showed, we're going to add more and more into the assistant. So, we can start with some very simple questions. So, maybe you've got a question on how to do something in our product. So, maybe how can I automate a reoccurring task in Central? And this will go out and look at our documentation and give that information and show you the sources. And that can seem really simple. A lot of AI agents we're used to. That's all it is. And then maybe it'll guide you to open a case if there's an issue. But for us, this is just the beginning of what we're showing and what's going to be available today. I'll note here we are in preview, but we are working on a, oh, this is actually very zippy. Oh, this is maybe the, we got it out very quickly. We are working to increase the speed here before we get out and about here. So, very quickly in plain language response, I have all the answers to what we're looking for here. So, how to go ahead and do that reoccurring task in Central. But we can also ask other information, so we can ask about it. What reports would show me device and hardware software details? So, again, it can go out and find that information. We can go beyond that. So, we can look at other information about our devices and be able to provide that information in plain language. And as we think about our help desk technicians coming in and wanting to find information very, very quickly, or business owners trying to, you know, maybe they don't use the tooling very often and they've got a very specific ask. Maybe they're on site doing a business briefing and something's going on and they want to directly interface. They may not be used to finding that information. So, we're finding that information. So, we're finding more and more that that's a lot of the pieces, not just using time, but also people that just aren't as familiar with the tooling. We got our answer here on where are the reports are and what's really important here is it shows the sources of the information and you can actually click right in and maybe get into the exact piece of documentation that you're looking for. It will go ahead and it will very quickly give you that response. So, again, that situation where you've got that panic call, you're on the news or have an alert about an active threat, how bad is your day going to be? How many devices have that threat? And being able to get in there very quickly and be able to see how bad that day is going to be. And so, what we'll see here is probably be patient. There we go. We only have a couple devices here. We have two devices in the environment. So, I know I only have to go find that devices. And actually, we can see right here that one of them was resolved by a patch or wherever we were. And it tells me about the information that's there as well. And I'm going to pop it back up because I have a feeling I know why that went wrong. So, let me try this piece here. And so, this is another example of trying to get information very, very, very, very quickly. I just got a call from the Dundee site that their logins are slow. Their server is called Avery. Can you see why their servers are running slow? So, again, very, very quickly trying to get information, trying to assess why is everybody panicking? What is going on? And trying to figure that out very quickly. You can see that could be a lot of clicks and a lot of issues in the moment. And very quickly. So, I do know what went wrong in that demo. There seems to be a blip in the vulnerability data. I probably should not have skipped ahead in that demo. So, back to the device area, though. So, we asked, hey, why are the logins slow? What's going on? And in very plain language, I've got an answer. So, this is a virtual machine. It only has four gigs installed, and it's running out of memory. And what's really nice is it gives me even more information on that. It shows me the performance that it has, which is for the last 24 hours, it has been running out of memory. There's CPU spikes, and it's giving me some recommended actions that people can perform. So, it can look at potentially increasing memory. It's a virtual machine. Check for memory leaks. Restart services if there's something going on. It's giving me both short-term advice and long-term advice on what I can do. So, again, the question was, I'm going to the site. It's really important that I know what's going on, so I know when I get in there, if there's any issues that I have. And in this case, it went and did an entire assessment of that site. So, it found issues that I need to be aware of. The server is out of memory. I have some patches I need to do. There's some updates I need to do and other information in a really good executive summary report very, very quickly. We've found that a lot of our preview accounts have really enjoyed this help and being able to get this reporting very, very quickly. Great. So, that is Enso, and it is really going to be really exciting, especially for a lot of our technicians and, again, business owners that want to participate in very plain language. I'm now going to move to the MCP server, which is another really exciting piece. And as Nicole talked in the introduction, this is where I'm going to show you in a cloud environment, the environment that I have. But if you start dreaming about the integrations, a real nice part about the MCP environment is anything that I'm doing here, you could do in an automated fashion. So, you can tie it into your ticketing system. I have it running in my team's environment on a side panel, so I can query the server directly in there. So, it gives a lot more capabilities. If you have other devices or other systems that have MCP capabilities, you can start bringing them together to do other automations or query across the different pieces that you have. So, that's really where the power is on this. So, again, in the MCP demo, very quickly, I can ask some very simple questions. Can you give me a list of all my computers? And it's going to go out and look for that for me. So, again, maybe this is an automated piece for your PSA tool that will go out and run that standard report. Maybe you want to do an inventory of devices every week for a client. Maybe if you have a lot of change, you could also ask the things, you know, tell me devices that have only come online in the last week or devices that have been offline for a while. So, I'll show you in a minute. But very quickly, I have this information. And that's just a simple list of questions or list of devices. I can ask for additional information. Can you tell me the user is for this device? So, again, you know, you could also, depending on the tool that you use, you could upload a CSV file and say, hey, here are all the devices for the users for this and it would quickly be able to do that for you and export it. And right away, this is my workstation and it's a workstation that it's on. Can you tell me about any Windows devices that do not have BitLocker enabled? This is really important, right? How many times have we had to respond to an information security questionnaire very quickly on all devices have BitLocker on or whatever the encryption solution is that you're using? Or you just, that's a standard that you have for your organization. You want to spot check every once in a while just to be sure that some of these items are happening. So, very quickly, in plain language, again, you can ask the information that you're looking for. And then going on to the PSA use case, you could also have this run on a scheduled job if you wanted to and maybe open ticket and assign it to your team. You know, anytime that there are devices that don't have BitLocker enabled automatically, you know, put an alert ticket out so that people know to do that. There we go. I've got a list of devices. A lot of these are virtual machines or lab devices that have been offline for a while. So, these I expect from the tests that I have. Speaking of offline, this is another one. I'm worried that I have devices that weren't turned in after people left. I think we all have dealt with that problem too. Can you tell me about devices that have been offline for more than 30 days? So, again, a very common use case and another thing that you might want to put into an automated action, automated report, or put it into a ticketing system for somebody to go find the devices and where they might be at a client site. And here we go. We've got a list of 14 devices that have been offline for more than 30 days, and I should probably go track these down and figure out where they are or remove them because they might be out of date. Now, again, we talked about the combination is bringing data together. This is a very simple example. That's just going to query the Web, but you could query documentation. You could query other systems. And so, in this case, I am worried about Windows devices that may not have enough memory or processor power during the operating system that they have. Can you find a devices that have less resources? So, the tool can go out. It can query. In this case, I know it's going to look at the Microsoft documentation. Go ahead and pull that. Tell me what it's doing and compare it against all the devices to tell me if there are any devices that are underpowered. So, starting to get in preventative measures or the other example I'll have is an app in a minute being able to plan. Do I have any devices? I can't run something that I want to roll out. There we are. Now, before my session, before I went in, I was looking at the questions before I started presenting and somebody said, I'm just glad that we haven't started with copilot. Here is that just for you. So, I want to roll out copilot desktop to all my devices. What devices do not have enough resources to run copilot? Desktop, but this could be any application, right? This is another request that we have. Somebody wants to roll something out very quickly. I want to go ahead and be sure that they can actually run it. So, if you think again, the automation use case, maybe before they go and request an application in their PSA tool, it could query their device and tell you if it's compatible or not being able to put those kind of automations in. So, they either need to request an application in their PSA tool. So, if you think again, the automation use case, maybe before they go and request an application in their PSA tool, it could query their device and tell you if it's compatible or not being able to put those kind of automations in. So, they either need to request an application in their PSA tool. So, they either need to request an application in their PSA tool. And in plain language again, I know what devices can't run it. Servers, we're not going to run copilot on a server, but there's one device here that does not have enough resources to run the application. So, if it's a critical application that we're going to be running, I know that right away I need to be able to do that. Okay. And the last prompt example I'm going to use, and I don't know, Nicole, in the questions we have any we want to show live, but I'm a manager at NMSP. Most people here manage the environment. Can you tell me what I should be worried about in this environment? Can you make it into a well formatted checklist? So, this is going to take a minute because Claude is going to go start an artifact and go ahead and start deploying it. But this is what you get, right? And again, you can pull other information and other pieces in to get this report. This is a report that will also show in ENSO. We had a little hiccup in that today and that's okay. Certainly, we'll be ready shortly and you'll be able to do this in directly in product. And it's a checklist. So, in this case, I could share it with the technician. If I had, again, PSA automation set up, I could say, hey, go make tickets for all this and do that extra layer on top of it. And I think as you start thinking about, depending on where you are in your AI journey, if you are going to adopt using the MCP server, deploying the client, think about how you want to deploy that environment. Think about what you're going to use. You keep in mind data security. So, we want to be sure that you're using tools that meet your data security needs. And that's something to be mindful of when you're doing that planning. And here we go. On the right, I've got this great artifact and I can go ahead and share it out. That's what Claude allows me to do. All right, that is what I have for a live demo. I love it. Thank you so much, Mike. We've got so many great questions and engagements. So, we're going to hop back to that in just a second. But in the meanwhile, let's tell everybody what we're seeing our end users see value in. All right. So, the customers who were kind enough to take part in our preview are seeing significant savings. So, if you take a look here, two hours to do diagnostics, that's what it would have taken before down to just minutes. And being able to look up documentation and best practices, as well as starting to look for things like freeing resources. So, stuff that's hard to do right now, but you can ask in plain English, which is great. In addition to that, as Mike pointed out, so we do have all of those MCP servers coming to life and you will have access to that immediately. If you go to developer.enable.com and you hit GraphQL, that's where you'll be able to see the documentation for every single piece of data that is accessible through an MCP server. So, developer.enable.com, and then there's a thing called GraphQL, and that is all of the data that is exposed for end site and end central up to your systems. And what makes it really powerful is when you're using an MCP server, as Mike said, you can connect to other systems as well. So, you can start to create workflows across all of your vendors. So, one of the most common workflows that we see our customers already doing is they take in or they get a ticket, that ticket has some information related to a device associated with it, and they're accessing the MCP server to automatically hydrate the ticket with the device information right away, so that they can then have a runbook kickoff if certain parameters exist. So, this is a very common programmatic use case that people are just using cloud or any other orchestrator to complete with our MCP servers and the MCP servers of our PSA partners. So, a lot of people have been asking, well, when am I going to get it? When does it get turned on? What comes out next? So, everybody will have access to opt in starting early next week. So, what you're going to see is there's going to be an Enzo button that appears up in the UI, but it will be a learn more until such time as an administrator in your environment clicks it and opts into the license. So, that's very important to us. We will not be forcing AI in anybody's environment without your consent. So, let's just, no worries there. What is being released next Tuesday for general preview, so this is preview, is the asset view. So, details about specific assets that is going to be available to you. So, you can ask a question about a device and get information, including diagnostic information back. Also, the product information. So, all of our manuals, all of our knowledge base articles, our support team has done an exceptional job over the last quarter, going through 15,000 knowledge base articles, getting rid of old out of date ones, making sure their new ones are up to date. And that is actually an agent that we've been using internally on our support team for over a year now. It's the enable integrated copilot MCP server. So, you're going to get faster answers about things that you would have had to go to support for before if it's available through a knowledge base article. So, that's already in there. And you're getting access to the external MCP servers. Super fun. What is coming next? Vulnerability. That is getting its own experts. So, you will have expertise within Enso to understand and ask questions about vulnerability data. Same thing with patch information. These will be coming out this quarter. Very exciting. As well as one of the most requested things is business experts. So, I'm going into a customer, tell me everything I need to know about their environment. What is their license utilization? That's going to be wrapped in a business expert. So, that's something even more than what you can just get with an MCP query because it has intelligence wrapped around it. The new script repo, if you're familiar with that, if you used that, that's going to have an Enso integration as well. So, they'd be able to generate scripts and to register them in with the script repo. And then what's coming soon thereafter, it's March later, but it's very high on our roadmap and we're starting to put it out, is your ability to specify your own experts and have that incorporated. So, your ability to put in your own standard operating procedures, to attach other MCP servers and have it in this well-controlled, well-governed environment. So, that's very exciting. And the ability to invoke scripts from there and to share this information across organizations. So, there's a lot coming. We're super excited about it. When you think about where we're going tomorrow, it's even more exciting. So, today, these are that selective AI agentic use cases. We will be coming up with cross-platform environments. So, if you are an Insight user, plus a COBE user, plus an AdLumen user, or an MSPN and even Passportal user, the ability to work with AI across our environments is top on our list. And it's top on our list to work with your other vendors. So, that's very exciting. And one thing that comes out right after that is you see all these AI agents and they're working across massive environments. The other thing that's happening is AI agents are shrinking down to a microscopic level. So, there is the ability to manage and understand endpoint devices at that microscopic level with AI agents. And that's when you start to get agentic autonomy and cyber defense on each device itself. So, that's that long-term vision there. All right. I'm going to go through. I think we've answered a bunch of the questions there. Mike, if you want to hop back on and we'll see if we can go through the questions one by one and hit them all up. Oh, my goodness. There's so many. All right. We're going to start at the top. Yes. Will Enso eventually be added to Automation Manager and help create automation policies? The answer is yes. Great question. Hi, Sebastian. It's nice to see you here. Insight, yes. MCP servers will be working with Insight. It is working against that backend. So, every piece of information available in Fusion is available through Insight and NCentral. Great call out on that. Thank you, Bill. A couple of questions on costs. So, right now, we are giving away the 100 communications. So, back and forth chat is what we call a token, essentially. And that is 100 per month per technician. We are looking at the added cost of if you need additional bundles. That will likely be priced, and we don't have a price set for that yet. But for the next quarter, it is anticipated that that is included in the cost of the product. Did you find a good one, Mike? I had a question that came in that I just answered. That's why I was typing the paragraph on. It was around running scripts from the MCP server against NCentral or Insight assets. And the answer is not today, but soon. It is getting close. We will launch you off to the right rights to do it. We are going to launch, because of preview feedback, a read-only endpoint as well. So, for some people, they just want that confidence that mutations aren't on. So, if you are in that category and want to ensure that, I don't know, either certain users or certain owners can't accidentally do something. It is not easy to do something, but we will have a read-only endpoint available as well based on feedback. Right now, there is not a separate MCP access cost. There are a couple of people asking that. Nicole, there are a couple of questions on data security in Enso. Can you talk about data residency? Yes. So, everything that we are doing is secure and it is isolated. So, we are not doing any automatic training on the language model. Some of the common questions that we get is, your data remains private to your organization. We are not storing your data. It cannot be accessed by anything else. Our language models are hosted in the private instance of AWS Bedrock that is associated with your current installation. So, it is geographically bound to the same location as your data is. We are not training our model using your data. So, there is a human in the loop. If you give us a thumbs down, because we have that typical thumbs up, thumbs down, and I have a few friends on here who are very good at thumbs down, so I expect to get a lot of feedback from you. If you give a thumbs down indicator on one of the answers, we actually have a human that is going back, looking at that information, abstracting the reason that it is a thumbs down, and without having any of your private data, there is an air gap, quite literally, between copying that information into our new test set. So, we are not doing any automatic training. There is no risk of contamination there. And your data is not shared with third parties. It will never be in use to improve an external system, and it is limited to all of our subprocessor statements now. So, if you go to our legal page, we actually have a full statement on our AI promise to you and our subprocessor promise. So, really careful about that. Questions coming in around Enso for on-prem and central. So, I think what we want to be clear here is if you have done the cloud opt-in, which is a simple way to say if you can see Asset View and see your data there, then you can use Enso. If you are truly on-prem, then you cannot. And the minimum requirements will be out there. You have to be on Enable Login as well. Yeah. We currently have a 19-day storage on your questions and answers for diagnostic purposes. For the preview, for the public preview that is happening right now, that information is secure. We can see that. What will be happening over the next month or two is you will also get a view into that data. So, for our super admins who have the ability to turn on or off the AI capability, you will be able to see the questions and answers for a 90-day period. And that will be controllable by the super admin. And to opt-in, everybody is going to get the Ask Enso button, and you just simply need somebody who is a super user to click it, and they will be brought to the license agreement screen. Nicole, what about questions around who can access? Is it on or off for everybody, or can you limit by the technician? Yeah. Oh, such a great question. For the public preview, which is what we're going into right now, the MSP, so your organization, agrees to it and turns it on, and it gets turned on for all technicians. So, it is not one individual technician right now. For public preview, it is everybody within your organization will have access. And just to be clear, Enso right now is read-only access. So, you will not have technicians that will be able to do things en masse or make updates. It is a read-only environment for this public preview, very specifically because we want to make sure that we get the capabilities right, and we get a high enough sentiment on the answers before we start doing anything that has actions involved. I know that a lot of our customers want us to move faster, but I am of the opinion that human in the loop is more important for experimental technology. There's a question on cost for the MCB server, and at this stage, there is no cost involved for you. So, you're welcome to use that, and it is in preview, and the skills that are there. In time, there may be some cost for really advanced actions, but at this time, there's no plan for that. Yes, the question, if you have a tech that has access to a single client, they cannot get access to other client information. It's quite literally inheriting the permissions of the logged in. There's a question on Copilot. So, we did use Cloud. We find Cloud has a higher degree of accuracy for providing answers, but if you are using Teams Copilot, Teams Copilot absolutely can connect to MCB servers, and you can choose the language model that it uses on the back end. Microsoft recently released the ability to use Sonnet within Copilot, and you'll find that that gives you a pretty good level of accuracy when you're doing MCB queries, so we recommend that. And if you're an Empower, Ben Lee and I will be showing that in the pre-day. Yes, if you're not an Empower, you're missing out, because that pre-day is pretty awesome. Okay. The AI tool. The AI tool is on for all users, remote control users, if they do not have access to see information related to devices. They will not see it through Enso, but they will have access to the knowledge-based capabilities. And that's actually, we're seeing significant time savings of people who just, hey, why isn't take control connecting, or why isn't something else happening? Our knowledge-based articles are actually really quite good on those, and it will give you diagnostic steps. So, they'll have access to that product information, but they will not have access to device information. All right. Moses asks if you can get a report on what your techs are asking Enso. Yes. So, the answer is it will not be available on Tuesday, but soon after, you will have a complete report on the questions and answers that are being asked by your techs. What model is Enso based on or using? So, we have a private Bedrock environment, and it is using a version of Cloud right now for that, the agent that's the orchestrator agent. As our experts get better and better, we find that sometimes using a different language model inside one of the sub-experts provides us with better results. So, we are doing optimization and A-B testing on each of those back-end experts, and we will be using different models, all privately hosted, all with inference turned off and learning turned off on each of those experts. Cool. I'm getting a bunch of questions on how to set up the MCP demo, so I'm going to go into the chat and put the link to the preview documentation and that is available today. That is a preview that's open to everybody as long as you meet the minimum requirements. So, that is the link to the GraphQL documentation. I'm sorry, the GraphQL MCP documentation. Nicole, there's a few questions on how to track usage, so with that 100 chat limit, is there a way to see the technicians that have used that usage yet? I imagine that's also getting ready to see how many people are using it. Yeah, so that will be available. That is not available next week. The technicians themselves will get indicators when they're at the 75% limit, so when they've used 75 out of the 100, and indicators as they go on, but this is a preview, so not all of that feature is available for the administrators to see, but it is coming. Yes, we've got a few questions that come in on linking your own LLM, so that is the demo that I showed with Claude, and as we talked about, you can do that with Gemini, or your tool of choice, Copilot, and so on. So, again, if you follow that preview link, you'll be able to follow the documentation to bring your own LLM, if you want to do the MCP server. Yeah, and Enso is not bring your own language model. Enso is a closed environment, so if you want to bring your own language model, you want to control the system, that is why you would use an orchestrator that's outside of Enable with the MCP servers. If you want something for your technicians to use inside the system, you would use Enso. We anticipate the majority of customers are going to use both. There are use cases that you're going to want to automate across vendors, and there are times when your technicians are just going to need a helping hand to move their inquiry forward in your environment, so you would probably want both. If you hit 100, Brandon, you will just be paused. So the ability to continue to consume tokens is paused at that point, and it resets on the first of the month. Enso does not work with patch reports today. That will be coming very soon. We will have it by the end of this quarter, and it will have all the patch information. It is 100 per tech, not a shared pool. Just to answer Jason's question, so you have a large number of techs that access your instance. You just get super lucky, and all of your techs get all of the tokens. So I hope they use them, and I hope we get a ton of good feedback. We got a question from Nathan. So Sentinel One information is not currently available to Enso. Just so if you're looking at Sentinel One information, you're going to want to do that with the MCP server to connect to both of those systems. We will have an expert coming out soon that will include that information. Yeah, oh, I love this. So John asks, will Enso have the ability to execute automated tasks on demand against endpoints? Yes, absolutely. It does not have that today, but we are bringing actions into the Enso interface, and that's coming out very quickly. So you will have the ability to do approved actions that are associated with the logged in user's permission, and it will make a recommendation for the action that you want to take, or they'll be able to call it on their own. That's the very exciting next stage. Great, I think we're starting to get through. Yeah, and Dave asks if Enso can see the ticket numbers of connected PSAs, so not in general preview, but that is an expert that we will be adding. Great question. Okay, yeah, and I do think Heath has a great, yes. I could go through this all day, but that's not what we should be doing. We're going to take a look through all of these questions, and we'll wrap up and give you a bit of an indicator of anything else that we should be doing, and I really appreciate every one of you taking the time to spend with us today to learn a little bit about Enso, learn about where we're going as an organization, learn about MCPs. I hope you all go out and adopt some MCP servers today, and on Tuesday when that Enso button shows up for anybody who is excited and interested, I hope you opt in and allow your technicians to see the benefits. It is preview, so please approach it with that preview mindset. The answers about the product, the answers about documentation, the answers about knowledge base, those are great, those are solid. When you start asking about specific assets, the specific asset information is good. A lot of people want to walk in and ask about all 1,000, 2,000, 22,000 assets that are under control. That is not part of this initial preview, so we end up with language model limitations when you try to drag in more than a million tokens worth of data, so I just want to be really cautious that the experts that we're releasing are the knowledge base and product expert and the individual asset expert, and the MCP that has access to all of the individual information related to that GraphQL. So thank you for your time. I truly appreciate it, and thank you, Mike, for coming on board with me and doing the demo. Thanks for having me, and we'll see everyone in power who can make it.
