Name: Stop the fire drills: Automate major incidents and reclaim millions
Uploaded: 2026-01-27T17:30:43.483Z
Duration: 44 min 36 s
Description: Stop the fire drills: Automate major incidents and reclaim millions

Transcript for "Stop the fire drills: Automate major incidents and reclaim millions": Good morning, everybody, or good evening, depending upon where you are joining us from. Very glad to have you guys tune in to this webinar, stop the fire drills, automate major incidents, and reclaim millions. We are seeing some folks joining in, so we'll give folks a minute or two, and then we'll get started. Thank you again for being here. Great. Let's go ahead and get started. So just in terms of today's agenda, we'll do, of course, introductions, but we'll spend time on the reality of major incidents, why they are becoming more frequent, more expensive, and harder to manage with traditional approaches. From there, we'll talk about what good looks like when major incident management, especially in environments that are highly distributed and fast moving. And then we'll walk through a live demo of BigPanda's AI incident assistant, showing how automation and agentic collaboration change the game during an active major incident. And finally, we will wrap up with practical guidance on how you can get started right away, whether that's evolving an existing process or piloting AgenTiC AI in incident response. We'll leave time for questions at the end, and feel free to drop them in the q and a, we'll do our best to answer them as we go along. With that being said, let's start off with introductions. Travis? Thanks, Manish. My name is Travis Carlson. I'm the senior product manager for AI products here at BigPanda. So I'm responsible for our AI incident assistant and AI incident prevention. Thanks, Travis. And my name is Manish Agarwal, and I work on the product marketing team here at BigPanda looking at how we can help with the agent tech IT ops platform for IT operations team, ITSM teams, and SRA teams. So, again, really excited to be here, and thank you everybody for tuning in. So we'll start with a simple reality check. Major incidents aren't rare events anymore. They are the norm. This slide shows the largest global outages just in the last year in 2025 based on down detector data. These aren't obscure companies or edge cases as you can see. These are some of the biggest companies out there, AWS, Cloudflare, YouTube, Google Cloud, Spotify, platforms that are used by millions of users and that invest heavily in reliability. And what's important here isn't just the scale of user impact but the frequency. Even the most sophisticated engineering organizations are dealing with major incidents on a regular basis. So the question isn't how do we prevent every outage because that's unrealistic. But the question is how quickly and effectively do we respond when incidents inevitably happen. And this just isn't a recent phenomenon. If you zoom out over the last decade, you see a consistent pattern, major outages hitting every year across industries, cloud providers, telecom, social platforms, SaaS, even security vendors. What's striking is that the, complexity keeps increasing, but incident response processes haven't evolved at the same pace. And as the slide calls out, we have barely begun 2026, and we already have had another big Verizon outage. And so the incidents are not slowing down. If anything, they are accelerating. And that's why the process you use to manage incidents matters just as much as the technology you build. And we when we hear from enterprises about these outages, the message is incredibly consistent. Outages are increasing in frequency and cost. There's a lot of metrics on this slide, but just to go over a few of them, over 80% of CIOs and CSOs say that outages are occurring more in the last two years, and many organizations are seeing 25 to 50% more outages year over year. And the cost to manage these outages and to respond to them has been increasing steadily. It stands now at around an average of $14,000 per minute. At the same time, leaders are recognizing that manual operations don't scale. And when leaders are looking to solutions, of course, AgenTik AI and IT ops are at the top of that list, and they are consistent consistently ranked as the key mitigation strategies. The takeaway here is important. This isn't just an ops problem. It's a board level cost and risk problem, and c suite executives and organizations are actively looking at better ways to respond. So now let's zoom in a little. Like, what does that major incident look like? This this must be familiar to many of you painfully. We often start by debating whether something is even a major incident. Then we page perhaps the wrong team. Once we realize that, we ask who else should we page? Who is the right team? And pretty soon, we are on that bridge call from hell with far too many people and teams that shouldn't even be there in the first place. We have the incident commander repeating status. The scribe is struggling to keep up. Teams are working in silos. People are digging for context across disconnected tools, asking, has this happened before? Do we have any learnings? So what's key here is this. Most of the time, the time isn't spent fixing the issue. It's spent coordinating humans and trying to figure out what's going on. So the workflows are fragmented, reactive, and heavily manual, and that's where the real inefficiency creeps in. So going back to that sort of manual process of solving these major incidents, when you step back and quantify that workflow, the costs become very clear. 50% of m t a TR is spent waiting for information. An average of 86 engineers let that figure sink in. 86 engineers on every bridge call, and it takes about seven hours just to complete post incident paperwork. And each of these has very real life impacts on how the organization is responding to these major incidents. In amidst this chaotic collaboration, there is unclear ownership. Because of so many people being involved, the institutional knowledge is lost, trapped in people's heads, and the slow post incident learning doesn't meaningfully improve future responses or the ability to, respond to major incidents. So organizations are paying a really high operational and financial price without actually getting better at incidents over time. But help is on the way, and so this is where BigPanda AI incident assistant comes in. At the foundation, the IT knowledge graph connects observability data, ITSM, change events, runbooks, collaboration tools, and even informal knowledge. And this goes across structured and unstructured data. And on top of that, AI incident assistant provides agentic capabilities, automatically assembling the right responders, swarming all of the observability data in your enterprise, surfacing historical context, driving root cause analysis, and automating communications and post incident reporting. And importantly, all of this is delivered through a conversational interface. Everything available in Slack and Teams in an intuitive UI as well as available in a web app for flexibility to meet users where they are working. And so responders of any experience level can get value immediately without learning yet another tool in the middle of an outage. Now I won't go into all of the details in the various sort of, you know, things that encompass AI incident assistant because you're going to just see all of that live. But a few of the things mentioned here are agentic troubleshooting, team assembly, automated comms, all synced across all channels. And so, again, this is something that will be really powerful as companies are looking to respond to major incidents. And this slide really brings that contrast home, the difference between manually attacking, major incidents and using AgenTeq AI. On the left, you see that manual reality, tasks that take anywhere from twenty minutes to two hours during an incident. And on the right, you see how those tasks are shortened to seconds literally with AI incident assistant, whether that is capturing historical insights, tracking change correlation, on call retrieval, managed incident channels, post incident reporting, and the list goes on. And the point isn't just speed. It's consistency. AI removes human bottlenecks from the process so teams can focus on resolution instead of coordination. And with that foundation in place, we can now show you what this actually looks like in practice. So, Travis, over to you. Great. Thanks so much, Manish. So what I'm gonna do, I'm gonna share my screen. We're gonna jump into a demo of this. And as we go through the demo, there's gonna be a couple things we'll see over the course of a a major incident that we'll we'll walk through together. And one, it's how Biggie will help to really automate that entire major incident management process, all the steps involved. You know, whenever there's a major incident, there's lots of communications that go out. And, usually, that's highly manual to craft those, to keep track of the bridge call, and put all that together. We're gonna show how Biggie helps automate all of that. And we'll also show how Biggie helps to turbocharge that investigation during the major incident, helping to surface key insights, as Manish was saying, to help your team quickly identify the issue, how they can resolve it, and get those services back up and working correctly. So let me go ahead and share my screen here. And what we're gonna see, we're gonna walk this through in Microsoft Teams today, but all this exists both in Microsoft Teams as well as in Slack. Alright. So let me share here. So you guys will see my Microsoft Teams now. And what we're gonna do, in this case, we're gonna say that we've maybe heard that there's issues happening, you know, maybe we're a a banking, organization. We've heard customers in our Florida region are telling us, hey. We can't access our ATMs. We're getting errors whenever we try to transact on the ATM network. And so in this case, I'm gonna go ahead and have Biggie help me start a major incident. And because, hey, Biggie. We've got a we've got a major incident. What's going on? So maybe it's kind of a keyword for it that it knows, okay. We we're trying to start a major incident. And when I do this, you know, I might have different processes I follow within my organization. A lot of our customers are large large enterprises, and they have different business units that might follow different processes for their major incidents. So you can have lots of different, we call them templates, but essentially lots of different kind of processes that different teams or groups might follow within your organization. And so this is gonna be one that I've set up previously, and all this is highly configurable. So you're gonna see my version that's set up, but you can decide, you know, are you using channels versus chats? Is it bridge calls as an existing call? Do you want us to spin up a new call? All these are things that you can choose. And so I just have to fill out some information and tell it, you know, a name for my my major incident. I'm gonna go ahead and give it things like, you know, emails, maybe who I wanted to send this to, and I can, again, I can have this prepopulated. You can see that in a few of these where I prepopulate it. You know, I'm gonna have it join a bridge call that I already have up and going. And maybe maybe my service desk already opened up a ServiceNow ticket, an ITSM ticket for me. So I wanna just keep using that ticket that the service desk opened. And when I click click create, now Biggie is going to begin this entire process. And so it's going to, first, in that channel that I specified, my my MIM bridge line one channel, it's actually created this banner message. Biggie's also joining my my bridge call. So you can see here, Biggie's joining that bridge call, it can transcribe what's happening in the bridge call in real time. You know, I can start putting in context. Maybe I can say, hey. A little bit about what's going on. And all of a sudden, we're seeing, you know, this banner message will update as people put information, whether that's in this channel like I just did, and it updates the banner. Or if people are talking on the bridge call, they'll hear that as well. And so let me go ahead and show, but one of the things I also have Biggie do as a part of my process is I say, hey. Whenever I start a major incident, I need you to send out notifications to lots of different channels. And as often happens in a major incident, you know, communications are really critical. When you're sending out to a broad audience, there's lots of executives involved. The way things are worded, the things that are said really do matter. So in this case, we actually you can choose to have every communication either sent automatically, or in my case, I've set this one up for human in the loop approval. So before Biggie will send this, it's actually messaged me directly and said, hey. Here's the here's kind of the notification I'm gonna send. Is this good? And, you know, maybe this looks good, but, hey. I'm gonna have to say, you know, remove the coordination note section and replace the open channel text with the actual channel name. And I can tell Biggie to kinda update this. So if something doesn't look exactly right, you can give Biggie a little bit of prodding to say, hey. Go ahead and tweak or adjust this as necessary. Now I'm happy with how this looks. And I can say, hey. Great. Go ahead and approve and post that everywhere. And it's gonna go do that, I've got it sending a message here. So it's sending a notification in this channel saying, hey. There's a major incident. You know, here's the the ticket, the bridge, as well as the channel for you to engage in in working this. You can actually see here, I've got my ServiceNow ticket, and Biggie's gone ahead and posted a note here as well saying same information. And I also have, an email that was sent out, and an email went ahead and reached out and said, hey. We've got a major incident that's going on. So, again, multiple channels, making sure everyone's aware that this is going on. Now if I come back in here, let's start working this incident. And as I work this incident, I'm just gonna keep providing information. You can actually see this transcription is updating as I'm speaking here on the call, with information, though none of that's super relevant to this. So it's it's ignoring it. And the banner message, you can see that. So one of the things that we worked really hard with our transcription is it focuses on IT operations specific information. And so as I put more messages in the channel, though, we'll see this continue to evolve as Biggie learns more about the incident. One of the things I'm gonna kick off is some agentic troubleshooting that Manish talked about. So I'm gonna ask Biggie to troubleshoot this incident. So it's gonna look at all the kinda information in the banner message, and now it's gonna go look across all the information and data it has to see relevant information that can help me try to troubleshoot and resolve this incident. In addition to that, I can ask Biggie to go check my observability. So I'm gonna say, hey. Go show me the health of my Florida ATFs. Check with my, you know, my observability tools and tell me if we see anything that's currently going on. And so this is gonna come back here in a moment. This will be the the output of that troubleshooting, and it's gonna tell me what it's found. And, again, it's looking across all of the organizational context I've given Biggie, and this can be historic ITSM tickets, changes, KB articles, Wiki pages. It could be call transcripts from previous major incidents or chat history from a previous major incident. All of that becomes relevant context that Biggie draws on to answer these sorts of questions. Well, that one's right now. I'm actually gonna click because I think, yeah, this one completed. This is where I asked Biggie to show me the health. And so it it validated, yeah, things are degraded. We're seeing in Datadog. We're seeing synthetic tests failing. In Splunk, we're seeing a bunch of logs, where we're seeing tunnels our VPN tunnels being torn down, so that's definitely not good. And anytime Biggie sends a response, it's gonna tell you things like the relevance as well as the data sources. So it's it's always trying to give you kinda transparency about where is this information coming from and how confident is Biggie in the quality of that information. Now I can also do additional things like, you know, when I start this, I can have Biggie automatically page relevant teams. But I can also, at any time, page additional teams if I need to get someone involved in the call just by clicking this button and selecting the team I wanna page here. Now let's see. This troubleshooting command came back. Let's take a look at this, and we see, you know, that you know, the first thing is saying, hey. Look for recent changes to your VPN. This is a common issue we've seen. And it gets into more detail. It says, hey. Here's all the times we've seen issues like this in the past and what the the issue or the root cause has been that we had to address. It tells us what teams and individuals are typically involved in resolving these types of issues. It talks about the priority. There's a there's a p one, which great. We've already declared that, so that's good. And it's giving a bit of insight into the the impact. And the final piece is really these diagnostic remediation steps. This is coming from places like resolution notes of tickets. It's coming from, runbooks or SOPs that it finds, about maybe VPN or IPSec, troubleshooting or or, things like that. And so we can see here, you know, a list of steps that I can read, you know, etcetera. And then, again, relevance as well as sources. Where did this content come from? And, you know, one of the things we saw, though, called out right away was changes. So I wanna just ask Biggie to help me out here and say, hey, Biggie. Can you show me any recent changes that could be affecting my Florida ATMs? Let's do a change correlation. And, again, often during a major incident, a lot of the customers we spoke with said, hey. The first thing we check is recent changes, and it's really painful. It's highly manual where there's people, you know, looking at previous changes in ServiceNow, trying to manually find things or looking for something with the same CI and hoping that they can, you know, find that needle in the haystack from the last twenty four hours, forty eight hours that might be causing this issue. And this is where we can use something like AI to quickly identify information. And so we actually found a recent change, and sure enough, we were updating our SSL certificates in that region. So we just pushed an update earlier today that updated those certificates. So it's highly likely that that's the issue. And, in fact, as the team starts to look at this, they might say, oh, yeah. Look. Hey. Biggie found this change. It looks like that's, you know, what's going on. And as we see discontinued updates, this incident kinda continues to evolve. And so we're gonna see things like information being populated up above. You know, here, we see that that change is now identified as the root cause. You know, it's got some more information about kind of what's happening here. And I wanna call out is not only do we have this information about the incident in the channel, we're also posting this in our web application. This is the the web application for our incident assistant, Biggie. And in it, we have a dashboard that shows you all the major incidents for your organization that might be going on at any given time. And if I click in here, I'm gonna see all the the kind of planned communications that Biggie's doing. So we see we actually have a status update that's about to run, in just a few minutes. I can also see all the incident details. So all that information I was seeing in the Slack channel and the banner message, it has it here. Because, again, you might have folks who are interested to understand what's happening, but they're not necessarily involved in working the incident. And so they can look at that information. They can see it here. They can see it through the comms that go out without having to bother the team who's focused on how do we resolve this incident and restore service. And, again, I can keep throwing in context, which will make for a nice summary when we when we get the the next automation here in just a minute or two. But as I put this information in, all this will get reflected in that update in the ticket. And, actually, you can see here, hey. I made a suggestion. And this is something we haven't talked about yet, but I I Susan said, hey. I'm gonna go roll back these old certificates. And one of the things that Biggie's gonna do is it's gonna look for times that you express intent to go do something. And when you express intent to go take an action, Biggie's gonna look to see, does it have relevant context inside of that big panda IT knowledge graph that could be useful or relevant to the thing you're trying to do? And so it's gonna let you know, hey. If you're gonna do a rollback to the certificates, hey. Make sure you're involved in the network team. You know, make sure you're checking with your your ingress load balancers and, you know, check with the service desk to make sure things have gotten better once it's done. So just kinda giving some insights from previous things that kind of happened within the organization. Now what we'll see here in just a second, we've hit that ten twenty three milestone, and so it's gonna post out this update via email and via I believe it's also gonna post in one of our channels of kind of the latest status. And so we'll be able to see kinda how big he's done that. We just saw that come in here to our major incident channel, and so we actually have an update. This is, essentially giving us a quick update of everything that's happening. It's posted to the channel. Similar to last time, I also had it send me an email. And so it's got a nice email update of what's going on and, again, links all the different bridge channels. And, again, if you think about for a lot of our customers, what we've heard is typically there's people or groups of people that are trying to do this as they manage, you know, leading the incident response. And so this takes that that work off of their plate so they can really focus on restoring service to the incident. And so what I'm gonna do right now, I'm just gonna go ahead and post a couple messages and close this out so we can see what it looks like on resolution because not only do we handle come back here to our MIM channel. Not only do things during the incident, but we also have this setup to generate the post incident communications. So I'm gonna go ahead and say, hey. We've got this resolved. I'm gonna resolve the incident and say, yep. This is this is all done. And so now Biggie's gonna kinda close out any of the tasks that are remaining, and it's gonna generate a nice kind of final postmortem or or summary of the incident. And, again, I've set up ahead of time where do I want that to be sent. Do I want to go to the the incident ticket? And, actually, that was one thing we didn't show, but when we posted that that recent update, the status update, it also posted as a work note in our ServiceNow ticket. So, again, as many places as you wanna send this information, you can have that set up to be done so that it's always posting that key information where it needs to go so the teams are aware of it. And you can see here now that this is resolved, here is the post incident analysis of what happened, and this is pulling from it's pulling from the channel where I've been chatting. It's pulling from that ITSM ticket and ServiceNow, and it's pulling from, the transcript. Now in this case, the transcript is me giving the demo, so there's not a lot of new information there. But if you know, a lot of times, some of the key nuggets of information during a major incident are things that are said on a bridge call, and they're not written down anywhere else. And so we're making sure that we capture all of that knowledge. And you can see here a really nice summary, a good timeline of kind of everything that happened over the course of this incident. And so we now have a really good sense of what happened, and we can start working with problem management and figuring out, you know, how do we resolve this and stop this from happening again so we don't repeat this incident in the future. And the other thing I'll call out is all of this data becomes context that's fed back into Biggie, and this becomes knowledge that Biggie has about your organization, so that the next time these things happen, or things like this might happen, Biggie can draw upon that knowledge, and it gets smarter and smarter over the course of time. So with that, let me pause here, and I wanna kinda switch back and just talk a little bit about kinda what did what did we just see. You know, we did a lot in about fifteen minutes there, but what we really looked at is, you know, how do we help, with our AI incident assistant to supercharge your incident management? Really, the goal here was how do we help drive smarter decisions across the team by giving them the right context they need, by using AI to look across all of the data that's available to you and quickly surface that insight and that knowledge when and where your teams need it. All of it was done in a way that was really focused on how do we make sure that you can trust the AI that's doing this. So we you saw throughout that there were things like human in the loop approvals so that humans still have control. There were also things like, you know, showing the sources, the relevance, anything that really provides transparency and clarity that's really critical. Even in some of the responses you might have seen, Biggie often will give its kind of logic of, hey. Here's what I did, or, hey. I went into ServiceNow, and I did these query filters to find this data. So it's really focused on providing as much transparency as possible. And and all that was so that we can deliver kinda consistent, efficient responses when you need them with the ultimate goal of reducing reducing the the overhead, the operational complexity for your teams while making sure that those key critical services are reliable for your end customers. And a couple things that we've learned over the course of this is we've worked with our customers and and done various research with them and and VBAs is things like understanding, hey. When when you have fewer outages, faster fixes, what does that what does it actually look like? You know, what are the metrics here that we've seen? We've seen MTTR reduced by anywhere from 20 to 80%. And actually, for for p ones, which are your critical issues, we've seen these resolved up to 20 times faster. So it's a significant impact in how quickly you resolve your most critical issues. We've also seen bridge call reduction. You know, Manish talked about it, is when you are able to page and get the right teams engaged from the start and you speed up their investigation, a, you have less people on those calls, and, b, the people that do join are able to more quickly identify the issues and get the root cause addressed so you restore service. And this is a really great quote from a senior director of observability at one of our financial services customers, and they said, you know, that there's been a huge reduction for them in major outages over the last couple of years, both in volume and duration. And this can really be directly attributed to BigPanda. And they said, you know, I've been at this game for over twenty years, and this is the first time that I've actually had real success in doing that. It's been kinda transformational for them as an organization. And that's the sort of impact that we're seeing with our customers. Now I wanna step back a second. And, really, everything that we've been talking about today has been focused on, our BigPanda AI incident assistant, but we actually have our entire AgenTic IT operations platform. And there's really three different products here. There's our AI detection and response. This is really focused on how do we help automate the l one response, that initial triaging of incidents, and doing this at scale. Right? Because if you think about major instances, there's a small number of these, but we're looking at everything. Your, you know, p ones through p fours. How do we do automation for that initial triaging of those running automatically running run books and figuring out where those tickets need to go. When those do need to go to an escalation to an l two or l three or it's a major incident, it can make a recommendation and hand that off to our AI incident system that we just saw, which is all focused around how we supercharge your major incident management response and how do we help kinda augment your SREs, your L two and L three. And then the third product is our AI incident prevention. And this is really focused on whereas the first two are about how do we respond when an incident occurs, AI incident prevention says, hey. How do we predict and prevent the incidents from happening in the first place? And this does things like looking at plan changes to identify what sorts of changes are high risk and flagging those ahead of time along with how you can mitigate risks so you avoid that major incident before it ever happens. We also do a large context analysis looking at problem discovery. So being able to identify by looking at thousands of incidents, hey. Where are these reoccurring problem areas? And making suggestions on how you can fix those problem areas to avoid those incidents from continuing to occur. So, really, it's it's really focused on prediction and prevention of incidents. And all of these different features are powered by our BigPanda IT knowledge graph. This is where we're pulling in data across, know, structured data, semi structured, unstructured data, things like ITSM service history, runbooks, knowledge based articles, chat history, all those things that have rich nuggets of information. And we've learned from a lot of our customers that a good portion of their knowledge is never written down in, like, a formal documented way. It exists inside of people's heads. It's in transcripts or in chat conversations. And so being able to unlock that information and make it available to everyone is transformational. So with that, I'm gonna pass things back to Manish to talk about how you can get started, with this. So let me stop my sharing here. Thank you, Travis. And, that was a great demo, and, hopefully, it gave everybody a sense of how, BigPanda incident AI incident assistant and Biggie in its conversational form can really help during those major incidents. And, as as we wrap up, we want to make this very practical. Adopting agentic IT ops or AI driven incident, management doesn't require a massive multiyear transformation. The most successful teams we have noticed start small, focused, know the metrics that they are going after, and outcome driven. The goal here isn't really to replace your existing tools or processes overnight. It's to remove the most painful manual steps and build confidence in automation over time so that it is not in your face. And what we see work best is starting with major incident response where the cost of delay, as we discussed, is the highest and the value of automation is immediately visible. And so this slide shows exactly where teams tend to start. So if you look at, on the left hand side, the AI incident assistant capabilities that automates the most time consuming parts of these major incidents. So standing up and managing incident channels and bridge calls, putting paging the right on call team so that we have the right folks on the call to solve the issue, automating summaries and postmortems for that for that quick sort of turnaround, and keeping all stakeholders in sync across the different teams and tools so that silos don't keep on forming. And what's important here is the shift. You move from manual coordination to orchestrated agentic collaboration. And instead of people chasing other teams, context, the system brings the right information, the right people, and the right actions together automatically, measured by outcomes like faster assembly, quicker engagement, and improved post incident reporting and learning. So what are the next steps here? First, connect with your BigPanda account team to see what proof of value could look like for your environment. They'll make sure to tailor it to your team structure, tools, and incident workflows. Second, take a hard look at your current incident management processes, especially as they relate to major incidents. Where are you losing time? Where is knowledge getting stuck? And where are the handoffs breaking down that are leading to those silos of knowledge being created? And finally, revisit how you think about operations overall. Incident response is often the best entry point into agentic IT operations because the impact is immediate and measurable based on those KPIs that we have been looking at. With that, I wanted to thank everybody for spending time with us today. We do see a few questions in the q and a. And, Travis, I think you're already getting to those, but it might be helpful even if wanna read them out for the benefit of everybody on the group. Yeah. That's that's a great idea. So let me just grab a couple of these. I've been typing to some, but I'll I'll kinda throw a few in. So there was one about what data sources has been Biggie been integrated with and what tools does it does it leverage. So when it comes to and this is when it comes to the data that Biggie uses, there's two types of data. There's data that we ingest and store in the BigPanda IT knowledge graph. With that, we have, kind of specific configurations for kinda some of the main data sources you might see. So I think Confluence, ServiceNow, Jira, etcetera. But we can also ingest really any format of data. So we have customers that typically you know, a lot of times, SOPs, and runbooks, you might have a formal date source of those, but you have lots of informal sources where people have Word docs, markdown files, PDFs, etcetera. We can index and ingest all those forms of data, and have with customers today. We also have real time integrations where Biggie might reach out in real time, and that's for things like ITSM data. So, again, ServiceNow Jira. It's also on call. So think, PagerDuty, JSM or Opsgenie, xMatters, and then observability tools. So we have a lot of tools here, Grafana, AppDynamics, Dynatrace. There's a I'll try to grab a link from our docs page. We have a long list of all these different data sources that we have today, and we're constantly adding additional ones, including things like supporting generic MCP servers, you know, lots of homegrown observability tools where, maybe we might not have an out of the box integration, but you can connect via an MCP server, things like that. There was also a question. can read up some of I, can read. off some of the questions, Travis, if that helps. So, we have a question that is asking, how do you add or create new templates? Yeah. So within let me I can just show this really quickly. Let me find even better. let. me I'm gonna stop the share. And I won't go through. a full full of the how the sausage is made, but I'll. just really quickly show you this. So within the web application, we have the MIM templates. And if you create a new template, you basically design this from scratch. So what are all the communication entities? And this could be again, I'm in a Teams environment, so it's gonna show everything Teams here. But, you know, bridge call, could be a Teams chat or meeting, a Teams channel, email, ServiceNow ticket. Again, I have ServiceNow enabled. This would be Jira if I had Jira enabled in this org, etcetera. And then you create all the things you wanna have happen. So, you know, the different types of actions that you wanna run on creation, onetime recurring on resolution. All this is something you build once, and then it gets leveraged every time you run that major incident. Sorry. for the picture picture there briefly. No. No. That's that's all good. I think you already answered this. So besides Teams and Slack, are there other forms of automatic state status communication supported? And you answered that, yes, we support email, ITSM, so such as ServiceNow and Jira. And soon, we'll also be adding a generic API that can be used to integrate with additional communication endpoints. No. So thank you for answering that. We have another question. Bridge calls are often a safety blanket for leadership during the crisis. So how has BigPanda seen successful organizations handle the cultural shift of telling executives, hey. We don't need a bridge call right now because. the automation is handling it? That so that's honestly been a big driver. And a lot of times, it's because we have the kinda automated communication updates, and you saw a couple examples of that. That helps provide because, again, what the executives really care about is they wanna know if something's happening, and they wanna get a sense of what that is and are we making progress. And so by having those automated updates going out, like, that's what they really, in a lot of cases, care to see is they wanna know, is this progressing? One interesting thing that we saw another customer of ours do, is they actually had a a bridge call for executives. And then the the team was working in the channel, but there's actually a separate bridge call that just the executives are on where they can ask questions. And then those questions get routed over to the team working, and then they, you know, they can answer them, it'll show up in the next update. So it was kind of a a way to give the executives the bridge call that they felt comfortable with, without distracting the team from the work they needed to do. Absolutely. Another question is asking that does AI incident assistant rank the information it shows by relevance to the issue and the sources that will be most helpful for the remediation procedure? Yeah. Great question. So, yeah, we do we do a lot of ranking in the background. We don't necessarily show that in the response of the relevance, but, yes, we do. Like, when we pull the data, it's it's ranked by those sorts of things, And then it's sent to the all in where it crafts the ultimate response. Great. Somebody was asking that, is Biggie GA? Absolutely. Yeah. AI incident assistant, everything that you saw is in full production, and it's available. So. absolutely. What data sources has Biggie been integrated with, and what tools does it use? I think we talked about that one, and I'll. I'll get to the the docs page that shows that. The other thing I saw, someone was asking about if you were talking in a call related to another incident, will Biggie kind of Yes. Yes. will it record one of the other questions as well. so. and this is, like, we have to be slightly careful here because, like, we're not gonna Biggie won't just, without a human prompting, go record calls in your company. That would create certain issues. So, but it will catch that reference. And so if you said, hey. There's another incident going on or somebody in chat referenced that incident and through the incident ID, yes, Biggie would catch that, and that would become part of, hey. This could be related to this other incident, x y z. And then it would be up to someone to say, hey. Let's go add Biggie to that as well so we can pull that in so that we still very much believe in a human in the loop approach with AI. No. Absolutely. I think it's it's like AI working with humans, and so that's that's the best outcome. And does this tool interface with paging systems in the event that the correct team members do not engage the incident management thread or channel? Yes. Yes. And so the way that we kick off the on call notifications, we we work directly with those paging tools. We have our integrations are built with those. So, like, if it goes to the first person, they don't respond. It'll follow the existing escalation policies that exist in your pager duty or your xMatters, etcetera. Yeah. Great. Somebody's asking, okay. We mentioned delivering value from day one, but how realistic is that? Like, how much historical data or training time, does the automated detection engine need before somebody can start trusting it to filter out the noise without accidentally suppressing something critical? Yeah. It's a great question. And, you know, a lot of the traditional techniques around AI would be using things like machine learning where there is, like, a training time. But what we're doing isn't that. What we're using is generative AI, and it's using a technique called, RAG or retrieval outcomes generation. So we index and store the data, and then we're pulling relevant data. And there's a lot of complexity that goes into it, so I'm simplifying here. But because of that, it's really just about ingesting that data, which typically will just take a few hours, and it's there and ready to go. We typically do recommend about a year's worth of data. If you think about when incidents occur, they might not occur, you know, but once a year, and you don't wanna lose that that kind of visibility. So at least one year of data is is pretty normal for most of our customers. We've had some give I think the most someone gave us seven years of data, which I would argue is maybe a little too much. But, you know, it depends on what you feel comfortable with as an organization. Great. I think you already answered this, Travis, but I'm gonna just call speak it out for the benefit of, all of the audience. Do you need to organize your support procedures and other reference information in a particular way to get the best performance and outputs from Vicky? Yeah. So I would say, like, there's no hard requirement on that. That said, obviously, like, the the more organized is always good. But most of our customers and and this is a common issue. There's some organization, but there's also a lot of decentralization where stuff isn't as always as organized as you might like. That's okay. We we again, most of the customers we work with have that issue, and it's we've seen Biggie perform very well. It's the sort of thing that AI performs actually very well in this sort of scenario. Awesome. Great. I don't think there are any other questions. If there are, please feel free to type it into the q and a panel. But, again, thank you everybody for tuning in. It was a pleasure having you. And as as we said, reach out to your account teams to do a POV, get started with the demo, and really see how AI incident assistant can supercharge your major incident management. Thank you again, and thank you once again from the entire BigPanda team here. Thanks, everyone. Thank you.