AI-powered features in Cisco Catalyst Center

I thank you for joining the webinar today. I'm Sumit Gugey. I'm the customer success lead from a Cisco perspective on networking experiences. Thank you for joining today's calling.00:00Is gonna be very much focused on, AI powered campus network with catalyst center, right? How can you gain more value using Catalyst Center specifically from an AI perspective and and how the team is doing a great job in building all the AI capabilities in the product now and in the future for you to.00:20To Gain more visibility faster at scale using AI. We have SMEs from the BE, Dan, Guyatrie Sandeep, Shay and team who's joined the call. So we're gonna record the session just FII team. So we will be sending the recording and the slides to all of you after the.00:40Call via an email. We encourage your questions. We are happy to take any questions that you have. If you can use the Q and A section on the right side bottom, there's a Q and A section out there, so if you can use that and we can answer all your questions while we are presenting, at the same time.01:00At the end of the call, we are gonna spend about 10 min for taking any live questions. So we will have dedicated time for questions as well. So with that, let's kick it off Dan, Grathry and team, thank you for joining.01:19Please go ahead. Thank you Samit, really appreciate it. I want to say thanks to everybody for joining today. It's great to have you all here. My name's Dan. I'm a part of the network platform product management team, and I have with me really the star of the show, which is my teammate Gayatri, who's gonna come in and do some live demos. I'm gonna kick the.01:35Things off today.01:55And what we're gonna be talking about is agentic ops and AI with a focus on how we're, we're continuing to add AI capabilities inside of catalyst center to help our customers become more efficient in how they operate.01:56Great their networks. I talked to a lot of customers in my role and there are some things that I and my colleagues hear consistently. The most common thing that I hear when I talk to IT executives is.02:16We're too busy. We don't have enough time to do the things we wanna do. We don't have enough people on our team, to take care of the network we, the way we want to take care of it. And that's, that's a really common problem. It's just not enough resources. And at the same time.02:32The stuff that they are supporting is becoming more and more complex. There's more devices connecting from different types of people in different types of places. So the whole landscape is becoming more and more complicated and.02:49AI, AI is changing everything, really, really quickly, very fast, and, and they're saying how do we, how do we adapt and make, make use of AI while avoiding the, the threats that can come with very fast moving emerging techniqu.03:07So as we're hearing these things from our customers, we're starting to.03:28Build out three core principles into our product portfolio. That is simplified operations powered by agentic ops in AI, security that's built into the network from the ground up and then devices that are scalable and ready to be part of the AI era.03:33We had a big week last week at Cisco. We were at Cisco Live and we made some really, really big announcements. I'm not gonna get into all of them, but one of the key things we announced was the launch of Cisco cloud cloud control. This is available.03:54In controlled availability in North America now, it is a cross domain AI native platform that is really built from the ground up by AI and for AI and agentic ops. I was rec.04:13Only talking to a customer, an IT executive at a large retail organization.04:33And he was saying, with all the different, tools and all the different data points, you know, being a, a a data driven IT organization, he said they have people that are spending more than half of their time just trying to correlate data across multiple platforms and and trying to, trying to make sense.04:39Out of, out of all this data from all these different places and wouldn't it be nice if they had a centralized place they could proactively help them do that? That is what we're trying to do with cloud control. Cisco cloud control is available to all of our customers. Entitlement comes from, entire.04:59Entitlement to the the devices you guys already have and to the the platforms that you already have access to. And it brings together information from all of the different services that you have from Cisco.05:19So security collab, campus networking, observability all comes together into one place for everybody. And just to give another view of that, Cisco Cloud control, you can see on the left of this slide all the different capabilit.05:37Abilities that are integrated from our software solutions including things like Cisco IQ.05:57And then what it gives you is centralized identity and and a single sign on experience. So you start at cloud control, you're gonna have your data from catalyst center, Meraki, Webex, all these different things as you drill down into those different solutions, you're not gonna have to relogin. Everything you log.06:02Login once and you're, you're there centralized inventory, topology of everything all the way from the client through the access layer, data center all the way to the cloud, all in one view.06:22Proactive alerts for you and then streamlined simplified licensing. So that's a big part of what we, what we announced today though what we wanna talk about is primarily AI and agentic ops inside of Catalyst Center. So we're gonna start with.06:35More with just kind of the network platforms, the way the way this all works is that the devices, your network devices, your switches, your wlcs, your APs.06:55They're sending data to our platforms. And then once we get all this data, it allows us to do a lot of interesting things around network automation, around assurance and visibility and then integrations common integrations across the platforms to help our customers do integrations with.07:06Other, software solutions that they have in their resources. And with all that data we're getting.07:26If we look at the combination of catalyst center plus Meraki deployments, we have by far the largest networking platform in the industry that includes over 40 million networking devices that are managed. You can see 27 million APs 8 million switches, so a huge.07:33Set of networking devices that are sending data to our cloud. Over a billion clients connecting to those networks every month. And most importantly, there's a lot of value in data and we're getting over 5 trillion data points every single day.07:54That allow us to do all these interesting things. With Agentech Ops and AI, there are a lot of things that we are looking for, for it to to help our customers with and really to, to.08:10Facilitate our customers delegating tasks to AI and to AI agents. Things like ticket resolution, how can it proactively help our customers not only respond to incidents that are resul.08:28By users, but really prevent that and and solve problems before the users get so fed up that they have to complain. Visibility and compliance as your network grows, how do you make sure, how do you know that your devices are running the way you want them to? How do you make sure that they're secure?08:47That they are, not running up against end of life milestones, things like that. And then really, a key thing that Catalos Center offers our customers is network automation, and AI is really gonna get us to the point where we can go from very many.09:05Manual processes to.09:25Automated processes to processes where if we have confidence in the system, we can allow AI to go ahead and make those changes for us based on what it knows about the network. And then just network optim optimization, performance optimization, there's a lot that it can do inside.09:26With catalyst center.09:46We have been working on AI capabilities for a long time. We were la we launched catalyst center way back in 2018 and started building AI into it with things like AI network analytics endpoint analytics. AI RRM is a big one that we have a lot of customer.09:48Customers who use that to help optimize their.10:08Their wireless networks, that's something that's available now for both catalyst center and Meraki customers common code, common foundations that that enable that to help our customers really make the wireless network work optimally for their users. And you can see we've.10:11Continued on all the way into this year where we are, adding to things like our event analytics, which goes through lots of messages to help find insights. And the latest and greatest thing which is really our showcase, which is our AI assistant, which.10:31I'm gonna talk a little bit about. So the AI assistant is, that's really a natural language tool that uses Gen AI to help our customers answer anything from simple questions like how to, how do I create a new SSID or how.10:50How do I, you know, make this configuration change or troubleshoot this issue? That's pretty simple stuff and we've trained it on our, you know, things like our documentation and Cisco's decades long history of, of having best in class understanding of networks.11:10But it can do more things like that. It can do things like answer questions like tell me what's wrong with my network. Where are there users that are having problems? Help me create an SSID, create create a network for me. Is my network ready for SDA?11:27There's lots and lots of things that it is capable of, and something that not everybody knows that I, that I wanna help get the word out on is that that AI assistant is available to everybody in cattle center 237.eleven.11:48So for those of you who are on that version or a or a newer version, it's there, it an administrator has to go in and enable it. It's not on by default. It's not a GA feature within the product yet. It will be soon, but it's available to everybody.12:07So have your catalyst center admins go in, turn it on, it can do a lot of things. Natural language and not just English but language of your choice. It's really good at doing multiple language to to expand on some of the coverage areas for the AI assistant.12:23Troubleshooting clients. Where are clients having problems? What are the problems? What's the remediation for the problems they're having, for our switching customers? Tell me about the ports, find misconfigured ports, troubleshooting connectivity, wireless.12:44There's a really common one. Wireless is obviously has its unique set of challenges, but it can really help with things like optimization for wireless, finding things like coverage holes and optimizing radio channels, running compliance.13:04Things like swim, tell me what devices I should be upgrading, where were there problems with upgrading devices? All kinds of things. So this slide has a lot on it obviously, but there's there's absolutely a lot that the, that the AI assistant is capable of. And I mentioned earlier that AI is moving really fast.13:22We're moving really fast with the AI assistant and we're gonna continue to iterate on it, expand its capabilities and add even more and more complicated capabilities to it to help it.13:42Answer questions for you that previously just were very, very difficult to answer. Moving on to other AI, AI ops capabilities inside of Catalyst Center.14:00The machine reasoning engine, this has been inside of call center for a long time. It's a really important set of capabilities. It helps automate complicated troubleshooting tasks. Think of taking a senior.14:14Networking engineer packaging up his brain, building it into catalyst centers, events, kind of infrastructure to help it, to help users solve problems and really solve problems for that. We've talked to customers who have said, hey, we NO longer need.14:31Your engineers, to work on these kinds of problems, we can have frontline help desks start to do the triage and resolution of problems because with one click, the MRE will go off and.14:51Troubleshoot problems report back on what's the cause and what's the recommended remediation for things like, network loops traditionally very difficult to solve, click a button, MRE can solve that for you automatically or flapping interfaces or all these different use cases.15:04It can help with a single click of a button run off and automatically troubleshoot, diagnose, and resolve problems for our users. I mentioned earlier AIRRM, one of the really common use cases inside.15:23Of Catalyst Center and now Meraki.15:43Wireless is really, really challenging and the bigger your network gets, the, the harder it is, the more neighbors you have, the more challenging it is. AIRRM uses lots and lots of data points from all the radios in your APs, and it does that over weeks of data. Analyzes the.15:46What behavior and predicts future behaviors and uses that to configure the right RF profiles, think things like what channels for the radios channel widths, power, all those kinds of settings it uses AI to make recommendations and automate.16:06The, the RF settings there to help optimize.16:26The RF performance. Finally, another key capability inside a catalyst center, AI network analytics. This really looks at the behavior of the endpoints, the devices that are connected into the network and looks for anomalies.16:31And and and and unexpected behaviors in those devices to help assess the risk and assign trust scores or threat scores to, to those devices and really help help network administrators know when there are devices that they they really.16:51I need to take a deeper look at, and, that's this slide here I was getting ahead of myself, but this is the AI analytics slide. Network analytics, my previous slide I'm going back to is more about visibility and optimization of of the network. So with that.17:10That's all I had for today. I want to go ahead and hand it over to Gayathri, who is gonna walk us through some of these capabilities inside of Catalyst Center. Gayathri over to you. Thank you, Dan, for getting us initiated into all things AI about Catalyst Center.17:30Let me begin with the most interesting topic of today's session are always on network agent, the AI assistant in catalyst center. What you're looking at is version 32X of catalyst center, the most recent one, which is FCS.17:51On the top right, you'll see a small icon here, which is your toggle to launch the AI assistant. So as I bring it up, I'm just gonna ask, begin asking a series of questions.18:10Dan gave us some insights about what intents are supported in our pilot release of AI assistant in catalyst center. Now be informed catalyst center is generally available starting version 23711 and version 32X which will be pretty soon generally available.18:26So, we saw a lot of intents both on the automation and network monitoring front that Dan flashed us. Let me begin by asking my assistant how many, ok, let's consider the scenario where most organizations have a software policy for their devices. So not.18:45They don't frequently upgrade their network device software unless they have.19:06I mean unless they meet a quarterly or biannual timeline criteria or there are critical security vulnerabilities that are mandating a software upgrade. Let us now assess my network to see how many devices we have those that are running either compromised software.19:12Or that are not running a software version that is recommended per your policy. I'm going to ask, I can simply, I can do this by navigating a few place pages here and there in catalyst center. But let me save some time here. I'm just gonna ask my assistant straight. How many devices in my network are.19:32Non compliant.19:52Can you gonna see how quick it correlates all this information? You might have seen some demos showing, excuse me, showing how you can access the network, the health of your network, monitor and troubleshoot, isolate root causes, root cause of issues in.19:55New network. Now we're also going to take a different angle, take a different perspective to see how we can use the AI assistant to achieve some automation tasks.20:15So there are 16 devices that are not compliant with my org policy one way or the other. How many of these are not running golden tagged image? That is an indicator that.20:24Either these are not meeting my org policy or they have a compromised software. You can see how it is able to, the AI assistant is able to correlate information between automation and assurance tasks.20:43Now there are eleven of them, which of these, let's ask, list these devices for me So I'm just asking it.20:58Plain English natural language questions, Nothing very technical, nothing sophisticated. It is able to summarize information from different pages or different features across catalyst center for me in a human, more human readable format that I can correlate.21:14Regardless of whether I'm trying to automate a particular task or I'm trying to isolate the root cause of an issue here. So once I have this list of devices that are not running, the recommended golden image, I'm gonna ask it.21:31Are any of these having sec, do any of these have critical security vulnerabilities? So I'm just evaluating the need for me to upgrade.21:45My software so it does not just come back with data, I mean, it does not throw back raw data. You can see data is easily tabulated in a more human read.22:05Correlatable format. It clearly summarizes in a tabulated fashion and it is limited obviously it has some limitation with respect to the number of show rows it can show here, but I go I can go ahead and ask it to provide me the complete list here. So here we have.22:26None of these eleven devices that are not running the golden tagged image have a critical security vulnerability, meaning I don't have an immediate pressing need to upgrade my software immediately now. That is our AI assistant in action. So in addition to monitoring the health of your net.22:43To work trying to troubleshoot and isolate a particular endpoint or user connectivity issue. You can also ask the AI assistant such questions to see how you can automate your network operations, day to day network operations.23:02Let me take you to the next, let me actually walk you through the AI memory lane of Catalyst Center. So just as Dan showed us, AI in Catalyst Center is not something that we've launched today with the AI assistant. Obviously that's our 1st step at the agent AI operations.23:17But as some of you know, we launched the machine reasoning engine way back in 2019. The machine reasoning engine is our capability that we built.23:37Based on at most advanced machine learning algorithms in Catalyst Center such that you can proactively identify issues in your network, assets your network, proactively to identify any issues that might come up a little later. Look at the impacted devices. See.23:49To what impact that this particular issue, if it is persistent in your network could cause.24:09What is the damage it could cost to your network? So we are here on the tools network reason now, as you can see there are a lot of predefined workflows that you can use right from assessing the CPU utilization. If, if you foresee any CPU utilization breached threshold.24:15If your.24:36Anticipating any interface down because of a maintenance window or whether it is overlooking a power supply, a spike. So there are a lot of predefined workflows that you can quickly launch to access issues in your network proactively. I want to run this vulnerability test just to demo how.24:38This works, when I run the MRE engine, what it does is.24:58I'm gonna have to re login into this instance again. So what it does is it, it performs a series of steps and comes back with its recommendations stating what, what are the impacted devices and what could possibly happen if you don't address these issues in the.25:03Immediate night near future.25:23So let's go back to that screen and take a look at what recommendations a machine reasoning engine has for us for that particular vulnerability.25:28So we started our AI journey way back in 2019 and oh we did continuous evolution by launching our machine reasoning engine.25:56Guided remediations for issues that we found which we populated using the device three 60 and the client three 60 pages. And then we evolved to some AI backed automation via the AI enabled.26:06AI enhanced RRM feature where which lets you define certain intelligent radio resource parameters allows you to run a simulation, a real time simulation to see how the modification to certain radio resources will impact your network. And then you can simply deploy that.26:22Configuration to your network regardless of whether you're moni you want to manage your wireless network using catalyst center or not.26:42We're gonna take a look at the AI RRM in a moment while I just complete showing you the network reasoner module here. So here's our vulnerability test. Alright, we'd rather take a look at the other.26:51Use case that we have, the AIRRM.27:11So we have a comprehensive workflow for the AI enhanced RRM workflow, which lets you both deploy your radio resource configuration.27:34To the wireless LAN to 9800 wireless LAN controller managed APs. Let's you modify certain RF parameters run a simulation, see how the performance is, and if you're satisfied with the performance, you can go ahead with the workflow, deploy the RF configuration on the wireless LAN controllers.27:45And after monitoring this configuration over this wireless network over a period of time, catalyst center comes back with certain insights about how these radio resources are impacting your wireless network. So we're gonna go to assurance.28:05On the health dashboard, you are, we have the AI analytics menu. So we'll begin with the automation part of the AI enhanced RRM. I'm gonna launch this dashboard. So the dashboard has two parts, one for the configuration deployment workflow, the other tab.28:22Here is to monitor the AI enabled sites where we have deployed these RF profiles. So the precursor to.28:41Launching and deploying the AI enhanced RRM workflow is going to be to define a wireless a AI RF profile. So under design network settings and wireless, you'll have options to create your RF profiles.28:51And now AI RF profile is just an enhancement to your regular RF profile with certain explicit radio resource settings that you can find only in the AI RF profile. So.29:09Settings such as dynamic channel assignment, the max, the flexible radio assignment, the dynamic bandwidth selection, the max bandwidth, the DBS max bandwidth, all of these are explicit to the AIRRM profiless profile. So this is a precursor that you disc.29:25Sign and create an AIRF profile before you go ahead and deploy it on an intended site or a building where you want to use AIRRM to dynamically monitor the really resources in real time. I'm gonna go back to.29:45Assurance AI RRM workflow. So when I launch this workflow, I'll be allowed to select the particular site or building on which I want to deploy the AIRF profile that we just created. So you, you have two options depending upon whether you want to continue.30:01Monitoring and managing your wireless LAN controller using Catalyst Center or not. So enable without device provisioning lets you configure the RF parameters without having to monitor and manage your WLC using Catalyst Center.30:22We'll go ahead with this option now. So there are certain other prerequisites. Obviously the 1st one is to have your AI analytics package enabled in catalyst center and to have the sites on which you want to deploy the AIRF profile assigned to.30:38I mean to have, to have the devices assigned to specific sites and obviously turn on telemetry for device controllability. And then I'll be allowed to select the building or site on which I want to deploy this configuration. I want to show you.30:58A site on which we've already deployed one of these profiles here. So on the RTP site here, the workflow asks you to define a primary wireless LAN controller and a different secondary wireless line controller, both of which have to be 9800 for the purpose of the workflow, and.31:17Subsequently through the next steps, you will be once you assign these wireless LAN controllers in these next steps, you will be allowed to select the AI profile which you want to deploy in the selected site.31:37So those configurations that we defined in our AIRF profile with respect to the FRA DCA and so on will get provisioned on these sites regardless of whether you're managing the 9800 controller using catalyst center or not. Now if I go back, once I have this deployed and I have.31:49I let Catalyst Center monitor this site for at least 3 h. I will then be able to see some insights, some meaningful insights that Catalyst Center.32:09Gives back in return. Now, we already have in our AI RF profile named RRM RF profile deployed on our London site. So here are some insights that catalyst center is showing for us for London. It says consider changing changing the.32:20Configuration for busy hours for RRM to be more optimal, and it also says expanding the channel list to reduce neighbor contention and to improve performance. So you can dynamically monitor the performance of radio parameters right on.32:38This dashboard. Now, when you want to tweak some of these parameters, you can run the simulation, change values on the RF profile, and then test it on the selected site. In my case London one.32:58So, Catalyst Center will run a simulation and show you what the changes in parameters look like, how they impact the radio performance of your specific site, in my case London one. And if your content with those changes, you can then ask catalyst center to either immediately deploy those changes.33:12Or couple it with an ITSM approval, have it routed to your ServiceNow so you can have your L two administrator review the changes and then push configuration to the wireless LAN controller. That is AI enhanced RRM.33:31From Catalyst Center for you. Next interesting thing that I'm gonna bring up is being able to baseline the KPI of certain parameters in your network. So, AI network analytics is our AI package in Catalyst Center. Once you have it enabled, you'll start seeing the site analytics.33:49I'm gonna begin showing you the site analytics. Now this is a dashboard that tells you the performance of a specific site over a period of time. I'm looking at a global site here over the past 24 h. There's a host of KPI that this dashboard monitors for me and lets you also drill down to specific.34:09Buildings or sites or buildings. It tells me how good or bad each of these KPIs are doing at each each each of the selected sites and lets me drill down to the specific time period at which a particular KPI is not doing good.34:29Now, what is more meaningful is it comes back with a series of insight. Now here in my case, it has listed me the top failure categories for onboarding and it also ranks the reason for, for these.34:49Failures in onboarding in the order. So when I say when I want to find out details about what has caused failures due to authentication category I can very well drill down into the details here. That's about site analytics. The next thing I want to bring up is be able to tell you how to.35:05Catalyst center's advanced ML algorithm can bring up baselines for KPIs in your network. So when you allow catalyst center to monitor your network over a period of time, it allows, I mean it generates baselines around how your network is doing for certain KPIs. So you.35:25We have a host of KPIs including onboarding time, failures.35:44And roaming failures here. You see a lot of B swarm charts charts here. The green dots are the green circles are baselines for specific locations. So for RTP one, we're looking at a 1.4 s onboarding time for four ssids with two endpoints connected.35:48Now, I also see some amber and red circles. So the the circles that are blue indicate the average baselines. Now anything that goes off is indicated by an amber and if catalyst centers MLL algorithm intelligencely is able.36:08To identify a deviation, it marks it with a red circle. So if I go here and.36:28Open further details, I'm able to understand there have been onboarding failures. Now, when did these onboarding, when are these onboarding failures reported, I'm able to drill down that particular time frame. I can I can also see what is a lower and upper threshold that catalyst center has baselined for onboarding.36:34Time here. So again, I can further drill down to see what are the impacted sites, what is the impact of this particular deviation, which are the clients that are impacted by this deviation? What is the root cause that is potentially causing this deviation?36:54So this is the level of detail that the advanced machine learning algorithm in catalyst center will let you drill down when you're monitoring the health of your network. Right from allowing you to baseline the KPI, it lets you identify if there is a deviation, what is the cause of the deviation?37:13When it happened and what are the imp what could be the potential impact of this deviation? That is about baselining. The next thing I wanted to show you is how catalyst center allows you to compare the performance of your network. When I say your network, you can compare different sites or.37:33Buildings of the same size potentially for specific KPIs. So I can select specific buildings here.37:53As you can see, I've selected London one and San Jose one for radio throughput band 2.4 ghz. Now here, catalyst center intelligently because it has a lot of data that is that it has observed and backed by the advanced machine learning algorithm, it is.38:01Able to tell me how the performance of these two sites compared for the selected KPI.38:21This is for the past week, right? So this is how we compare the performance of two different sites. We can again do this for two different device families as well.38:29So you can go and compare the performance of your network peers per specific again for specific KPIs for certain time period over the past week and we can ask it to highlight how the peer.38:45Performance has been. Now, what is indicated in blue is the performance of your network versus what is in puble is the performance of your peers of the same size almost. So right from allowing you to automate your network using a.39:01Intelligent AI RF profiles to allowing you to deploy it, to optimally use your radio resources.39:22Two, allowing you to baseline your KPI for network health monitoring, comparing it between different sites or different geographies in your network and comparing the overall performance of your network with your peers network. There's a whole lot of information.39:30Correlation that catalyst centers advanced ML algorithm can do for you. And all of this is in more readable format than just numbers here. Let's take a look at the trends and insights now. Another interesting or sophisticated feature I'd say.39:50From catalyst center is its ability to identify if there are potential potentially degradable user experience cost for your clients with specific APs. Now, if there are APs that have higher client activity which are causing downgraded user experience.40:08Experience for some of your clients in your probably in your campus network. Catalyst center can bring up and highlight it for you.40:28So here we've selected 2.4 ghz radio band and we see there are five impacted sorry 20 impacted endpoints because of height client activity. Let's take a drill down into this. I'm looking at traffic distribution here.40:37For 2.4 and utilization distribution. So, I can drill down into further details as to what are the three radios that are causing bad experience, what are the reference radios that catalyst centers picking up to say these radios are causing bad client.40:57Experience. Now it does not stop there. It goes ahead and tells me what are the impacted clients because of this high client.41:17What are the endpoints that are facing potential bad potentially bad user experience because of this one? So right from being able to baseline KPIs to telling you how your network compares with peers and how different sites within your campa.41:29Those network are performing against each other to showing you a trend of how the client experience has been in your network to identifying if there are potential problems in your network. There's a host of things that catalyst sent us.41:49Machine learning and advanced machine learning algorithms can do for you. This is pretty much what we wanted to show you today. I'll go back to Sumit and see if we can answer the open questions that we have.42:06Yeah, so there are a lot of questions, great, great interaction in the chat and in the Q and A section. So thank you team for asking and thank you Sandeep Dan team for answering. One of the things I'm gonna open up the.42:26I'm gonna open up the, the audio and unmute everybody to ask live questions, but before that, I have a few that we have captured that is it like a common theme of questions that are coming up Dan and three one is.42:42I think one question is, ok, what version do I need to be on to be on catalyst or to be on catalyst center for using these capabilities? I think we've answered that. Anything greater than 23711, is the version. But the specific question that came in was.42:58I am on 23711, but I don't see the option. How do I know, that's a good one. So for 23711 is we've established a process where you reach tag and we enable, AI assistant upon request. So the the specific catalyst.43:14In the release on which catalyst and the AI assistant will be by default enabled will be 32X.43:34Okay, thank you. And 32X or the the three.train GA, so there was a question about that so I think the GA is tentatively planned end of June. We'll of course let the team know when it gets confirmed.43:40But that's the timeframe, correct? Right. I I'm gonna just maybe counter that a little bit. Initially when we launched the AI assistant in catalyst center, it, it, we had to be involved in updating it.43:55But my understanding is that we've made it available to everybody, that a super admin user can go in. There'll be a banner message when they 1st log in, if prompting them say, hey, AI assistance available if you want to enable it, go here and they're just a click or two away from enabling it. No Cisco intervention needed.44:12That's my understanding.44:32Sure so just to kind of confirm on that is if a user is not a super admin, is this capability not available for all the users who are using catalyst center like network operations team etcetera? Yeah. So the, the answer is that once the AI assistant has been enabled in catalyst center, it's available to.44:35Do all catalyst center users. There's NO, there's NO way to break some users have it, some don't. If it's on every catalyst center user has access.44:55And at least for now, we may, we may, expand on our role based access privileges sometime in the future, but in the initial releases, it's all or nothing. Sure. There was a question about a lot of questions came in about.45:05Can AI RRM or or all the AI capabilities, is there a way for doing, getting email alerting? Is there a capability of building custom reports and getting custom reports from AI analytics et cetera?45:23Oh, that's a, that's an interesting question that's kind of a broad question. I can say this that the the AI assistant turns out to be really good at.45:39Kind of custom reporting types of challenges. I want to report on this type of users or this, these, types of devices or things like that. It can do that really well. It can use, you know, graphical output.45:54To do those things. So that's one way, one new AI based capability that, that can definitely help with those kinds of things. And definitely within catalyst center, there are, the, there's a pretty expansive reporting framework. There's lots and lots of different types of kind of pre canned reports that are.46:11Available. So have a look at those and see if any of those work for what you're looking for.46:31Oh Chuck. I was gonna say there's another one that just came in asking about the AI assistant and whether it works with SDA and fabric deployments, the answer is absolutely.46:38Yes, it works for all kinds of deployments, all kinds of architectures, topologies, it can help just about everybody.46:58Team, you can also raise your hand that I can see here. So there's an ability for you to raise your hand to ask a question and then I can go ahead and unmute you. So you can also do that.47:14There are a few more questions that came in. Yeah, there there's a good one. Frank just ask a question that that I want to take because we were just talking about the AI assistant and role based access.47:25So what I had described earlier is that if the AI assistant is enabled, all users will have access to the AI assistant, but if they start so if they have a, a catalyst center user role that has limited.47:43Functionality they they they're not gonna get, they're they're not gonna be able to answer questions about stuff they wouldn't have access to. So they don't get, they don't get privileges to more information from the from the from the AI assistant.48:01It still respects the existing role based access. They'll have access to AI assistant.48:21But what AI assistant responds to them with is only the stuff that they are entitled to have access to. And all of these features are available with the existing catalyst center, there are NO specific licenses required.48:25For utilizing AI assistant, utilizing AIRRM and point analytics and all the features that we spoke about today. On AI, all of these are available for all center customers. Well, so there's.48:42There's a couple of considerations. One is that the vast majority of our AI capabilities that the AI computations actually happen in our cloud. So the catalyst center needs to have, access to the cloud.48:57If you're a customer that's running in a, a pure air gapped environment, catalyst center has NO access to the cloud at all, then those capabilities won't be available.49:17And in general, there we have multiple licensing tiers. The AI capabilities for the most part are available for customers in the advantage tier.49:30There's a question about will CCGM have AI analytics dashboard, have plans to integrate four of my clusters? That is a good question. I, I don't have the answer off the top of my head. I don't know if Gayathri or any of my.49:50Colleagues have an answer for that.50:08That that's something we might potentially have in a future release of CCGM, but I mean what we plan to also bring about is the AI assistant for the CCGM when you have integrated clusters, that makes it way easier for you to analyze and troubleshoot issues.50:11Right. There are multiple questions, so we will be sending out the recording and the slides to everybody via the email that you used to register for this webinar. So we will be sending that. I see specific questions about having AI network analytics enabled but not seeing trends data or baselines data. I suggest.50:33Or request you to directly reach out to us.50:53So we can evaluate your deployments and provide appropriate suggestions. Right, I see I see a question about what is the implication if ICE is not integrated?50:57Into catalyst center. Catherine, you wanna take that one? Okay, the requirement of having ICE integration is for external authentication or AAA authentication. So when you have your users locally authenticated, there will, you might see a lot of limitations in.51:16How.51:36Your endpoint monitoring happens There's a question about we have NO Cisco wireless controllers and access points.51:36Our enterprises using Aruba?51:56How can I get started? Unfortunately, we have not extended the AI features to non Cisco hardware yet. So for any non Cisco gear that you're managing using catalyst center, you will only have limited visibility under inventory and topology.52:00All those information that is published via RFC 1413 is what you'll see in Catalyst Center. So, sure. Yeah. And the AI but the A if you're if you're using catalyst center for your switches and routers, then obviously it will be available.52:19To help you solve solve problems and, and.52:39Do work well for those types of devices Will the AI assistant work with SDA and fabric deployments?52:43Yes, absolutely yes, they will alright we have also unmuted everybody, so if anybody wants to go ahead and ask a question, please go ahead or raise your hand.53:02Alright, if there are NO questions, thank you.53:47Everyone for joining. There's gonna be a quick, quick survey after you close the call. There'll be a quick five question survey, really appreciate if you could provide us some feedback about how the session was helpful, if there's something we can work towards adding more value to your network management platforms, any features, capabilities.53:51That you think are missing today that that are important for you. The quick survey is gonna help you do that and we will anyway be sending the slides and the recording to the email that you've registered with. All right, thank you team. Have a good day.54:10