all right well we're two minutes in so we're going to get going uh as I said before good morning good evening good afternoon depending on which part of the world you're in uh my name is Shashank Sager I'm a senior director with the itm business unit and I'm really excited to talk to you together with my esteemed colleague Jason leang about now assist um features that we've just launched in the November release so before we jump in uh quick Safe Harbor a lot of the the information that you're going to see here is related to the to the member release but some of the questions that we'll be answering for you uh might be related to some of the forward looking areas that we're still working on so just be aware that these are not meant to be shared with customer as is uh because there's a likelihood of some of this information to be able to change also now uh before we get started live on service now uh you all have probably attended it so this is an interactive event series there's a barcod there so if you're interested in attending any of the future events uh we promise that we'll have many more of these coming for now assist you can scan this barcode and voila you'll be in to be able to have a full schedule and then you can sign up for whichever ones you want to be able to see a few housekeeping items today um we did save time at the end for Q&A uh but meanwhile if you have more questions that come up you can always use the Q&A button at the bottom of your screen along the way and depending on if Jason or me whoever is not presenting will be sure to keep up with answering the questions if you're not able to answer the questions here uh we'll also make sure that we take some of those questions and we answer it towards the end also this presentation will be recorded uh and it'll be shared on the service now community so you can have a look at it later if you want to be able to refer to it as well and you can always reach out to Jason and me for any questions that you might have also now after the event you'll also be prompted to fill out a short survey so proactively we thank you for the feedback feedback is a gift and uh any feedback we get is always used to be able to make these sessions more interactive for you as well as U um valuable so we can make the best use of your time now without further Ado I I introduced myself uh already but I'll give a little bit more colors so Shashank Sager I've now been with itsm Buu and service now for about two years and prior to joining service now I have about more than a decade of experience with B2B SAS businesses with larger organizations like Adobe as well Jason yeah thanks Shashank my name is Jason leang as Shashank mentioned earlier I'm part of the itm product management team been here almost four years joined during the covid time frame um previously I was had a startup focused on Virtual agent nlu type Technologies so as um as I've been focused previously on our virtual agent and AI Technologies I've moved more toward uh some of our geni focused capabilities um nice to meet everyone looking forward to sharing what we have for our November release and moving forward fantastic yeah Jason's definitely my partner in crime with with now assist and what we're bringing to the market uh clearly since 2022 our lives were turned upside down with Chad GPT being launched uh and this Market literally changes by the day so really exciting times but at the same time super interesting to see where this Market moving uh technology is going to take us as well now before we jump in uh we wanted to do a quick poll um so Jason are you able to fire it up yeah let me go ahead and kick it off here and really the question is um so we're going to give a two minutes for everybody to be able to answer these question we want to be able to know and understand what's the timeline for now assist for customers uh or if you have any plans do you plan to be able to activate it in the next quarter next year you don't have any plan to implement it yet um like I said said if you if you're not a customer let's say if you're a partner if you have a customer that you work with what are their plans and a lot of this information will help us know and understand uh what's the overall uh Readiness for this product and solution in Market uh but we also use this information as a way to be able to uh determine which of the new features we want to be able to bring to Market as well yeah and for folks that answer no plans yet to implement would be great to just understand and um you know what what are some of the reasons maybe you aren't um super familiar with the capabilities feel free to drop it in the chat that way it'll give us some additional um context as well that's a good point Jason yeah maybe there are certain features that you'd like to be able to see as a part of now assist that that we could think about prioritizing for a roadmap so now is the time to bring all ideas okay so um there's one actually answer that was that not enough awareness about this capability and feature so hopefully we can provide some of that today everybody 30 seconds more una aware of this capability okay you just see see what it's capable of before planning and implementation the good news is that the implementation is super seamless so Jason's actually going to show you a demo of how easy it is to uh set up virtual agent actually with NIS today licensing requirement it's a good question we'll cover that too okay I think we're good let's got about 75% go ahead and share the results of the poll right so we've got about U 5% implementation in the next quarter 42% implementation in the next year and no plans to implement yet about 53% so keep your answers coming if you don't have any plans to implement or if you have plans to implement next year but if there are any barriers that you believe we can help uh solve so to help accelerate that timeline so keep the ideas coming great all right well let's then jump in to to um give you a little bit of a taste of what's coming in November but before we jump in just um a little bit of a detail of how service now has uh thought about what generative AI means for our customers and and really the way we think about it is that we want to be able to increase efficiency while maintaining exceptional user experiences but in the context of what are some of the more feable task that could be automated using generative AI so if you look at some of the the market data that's available from Gartner McKenzie or Harvard Business Review there's a lot of data that actually has indicated to us that there are these tasks that are there for each businesses whether it means that the time spent or or time lost to reorient it to task switching or whether it is around time loss searching for information to complete a a task for a workflow well I call it the swivel chair uh but what it does for customers at the end of the day is if let's say an agent is trying to resolve an issue and they're not able to get the information in the shortest amount of time then it does cause a little bit of a um uh rupture of the customer experience overall because now the customer has to sometimes share the same information twice with the agent or in some cases the agent takes longer to be able to get to resolve an issue potentially so some of these areas were the one that we were very early in the day looking at and uh the approach that service now has continued to take is while we want to be able to develop or bring new features for now assist that are really aligned with our expertise around applications and workflows we want to be able to do it in a way so that they're really easy to set up and most importantly they're really easy to use and they drive productivity for different employees in the organization and there are three primary uh employees that we think about in any business one moving from left to right is a service owners uh so the admins that are responsible for setup how do we make them more productive uh we have the agents and then we have the employees so for each of these personas we're trying to dig really really deep into understanding their pain points their needs and then be able to prioritize features which we believe will get them the biggest value for their bu so for an agent it might be something as simple as uh they just got an incident but now they want to be able to have chat summarization available for them to be able to resolve the incident faster increasing the agent productivity but for employees it might be about how does the overall employee experience improve because now the agent is able to resolve their incident faster with less amount of information that they have to give because that information can already be derived from the system itself a lot of these are based on the principles of KCs Knowledge Center support um which which are best practices that we often see that the customers want to be able to deploy for their Incident Management best practices as well our goal with now assist is not only to be able to introduce a few features but over a period of time introduce features that span across our entire portfolio So today we're going to Showcase a few functionalities that were launched as a part of November release uh specifically related to chat summarization resolution notes as well as virtual agent but for those of you who were with us in the previous live on service now session where we uh showcase some of the Vancouver features uh there was much much more and I can tell you working very closely with Jason and the rest of the product teams there's so much more coming so this is only the beginning of but at the same time super critical and important for us to hear from each of you on what would you like us to prioritize so keep the ideas coming keep the feedback coming in the chat b as we go along as well uh uh one of the customers who we actually worked very closely with when we launched now assist was corera U and I personally worked with Darla at corsera she's a great partner so we wanted to share with you some of the early U quotes and and um verbage that we're getting from Darla and businesses like warera we're already starting to see the impact of some of these functionalities and we also continue to do a lot of internal studies so to share a data point one of the the data points we've recently continued to monitor is that agent productivity agents who actually use Nows this functionality their productivity increases anywhere between 12 to 20 U over a full day now obviously that translates into a certain amount of dollars but these are internal to our own teams for every customer this means something completely different before we jump into actually showing you the actual functionality on how some of these features are used we also wanted to take a little bit of time to walk you through how does now assist help Elevate the overall experience both on the employee and the agent side these are the two personas that we're starting uh to focus with uh but we're also going to rapidly expand our focus and attention into the admin portfolio owner uh Persona as well but imagine a day where you have an employee she logs in to the system fires up her computer um and she's looking for some information around parental leave so she goes to um the employee Center or the internal portal to be able to look for some knowledge articles uh unfortunately she's not able to find any knowledge article in that case she fires up virtual agent bot to be able to self service be able to find that information herself that too to no aail she's not able to find any information strike two her last resort is now to be able to talk to a live agent which she requests through the virtual agent conversation for an Handover this is typically where in the current life scenario uh there would be a triage conversation that would happen so the agent level one tier one agent would typically try to get the information from the end user to try to understand what the issue is they would try to find knowledge articles to be able to find answers to those questions and in some cases also try to find similar incidents to see if there's anything documented around how to resolve it often times uh the the tier one level one agent is not able to find a resolution so they reassign the incidents to a level two tier 2 agent who in this case might work with a subject matter expert to be able to resolve the incident uh and then they would end up creating a knowledge article that can be then used in the future by level one or even other level two agents to resolve similar incidents as well this is where now assist comes in so the way we Elevate our overarching features is that number one chat summarization so chat summarization is the one where during the triage conversation chat summarization is able to provide a really quick summary back to the agent so the broken experience of having to ask the customer twice for the same information doesn't exist anymore and the agent can continue the conversation from where the virtual agent bought left it reassigning the incidents this is also where we're able to activate now assist uh this is where we' be able to provide a summary of what the steps have been taken so far so when the reassignment happens to level two level two agents now starts with a rich way of information of what happened not only on the level one agent conversation but even prior to that on the virtual agent bot conversation summarized in a really short succin way now keep in mind whatever I'm showing here is all powered by now now llm so we now service now has its own own power llm that is a proprietary llm that we're managing for all of its customer as a part of the the features that we bring show here they are available to you as a part of a Pro Plus or an Enterprise Plus Q so if there's a desire anyone to be able to uh purchase it please work with your account manager for U pricing and any other details I do know there are some really nice promotions that are ongoing for Q4 so if anyone interested now would be a good time um the next part is around resolution of incident so when an incident is resolved um how do we provide a quick way for agents to be able to have a summary of the overall incident so we have that available in notes and then finding similar incidents has a similar capability as well of resolution notes generation and then last but by no means the least uh we have the ability to generate KB this functionality is not live yet but we're targeting something around the the March time frame what we are going to Showcase here today though uh Jason is going to show a live demo is around virtual agent conversations within um using the new now Assist Technology so those are all the features that have been uh we've been able to bring to Market since the Vancouver and the November time frame and really what all of these uh use cases do to be able to improve the overarching value metrics so help accelerate uh meantime to resolution as an example A lot of these areas specifically focus on increasing the agent productivity which means that the time it takes to read an incidence or time to take it take to edit the summary is reduced overall making them more productive um also from an employee experience perspective as the it takes less time to be able to resolve an incident there's more um self-service there as well as well as less number of escalations the employee satisfaction and their overall experience increases also so this is what we call as the main anchors for now assist that we have any questions so far before I jump in and hand it over to Jason for what what's coming in November yeah there's just a quick question from the Q&A is this llm the same for every customer or would there be some changes for customers um today currently it is a a central now llm model um it does leverage um information on your particular instances but the model itself is is pre-trained by us currently that that's the uh only model uh in the future there may be some additional um fine-tune models that might be available but that's the current implementation uh that's all I see any other questions before we kind of move on okay so I'll move on Jason I'm gonna I can share the first few slides and then I'm sure you want to be able to share your screen for the demo yeah yeah if you just share the first couple slides we can we can start with um talking about some of the capabilities um I know everyone's kind of coming from this uh a little bit different differently in terms of understanding the capabilities sh Sean kind of laid out a little bit of what we initially started with um which is around the summarization and generation type use cases which traditionally when you look at large language models um they're very good at creating content generating it and summarizing it so when we look at the initial flows we've been initially focused on incident so a lot of the experiences we we we launched in September and August time frame and then we built and enhanced some of those capabilities in the November time frame so one of the areas that we looked at is previously um and I might be able to show this if we have some time but previously a lot of the summarization capabilities required um agents to manually go somewhere click click a button to be able to summarize the particular record um what we found is in in talking to a lot of Agents is um this kind of distracts a little bit from their existing workflows so how do we kind of bring some of those capabilities in line into their um current experiences and how they're working on incidents so when we look at um a lot of the summarization type touch points usually you look at things like during agent handoffs reassignment type flows um coming in during some sort of chat conversation so those are the areas we started kind of really looking at and and so when we're talking about this slide enhancing the in workflow summarization experience we want to kind of make sure that this becomes a very natural intuitive um experience so that it it really just augments how they're using these capabilities versus changing their mental model of how they're addressing tickets so really in this case one of the areas that we focused on was like I mentioned the handoff or assignment so often times when you have potentially either an level one escalating to a level two two or maybe a level two that has been misassigned a ticket and needs to assign it to another group uh oftentimes you have to kind of go in and uh type in some background on what what has happened already on the incident uh what the communications have been potentially if they're not in the notes so this gives opportunity during reassignments to um generate a summary based on information from the today it's from the short description description and activity streams between the work notes and comments um this in the future will be customizable in terms of the input today those are kind of the core Fields we're looking at and it also does if you kind of look at this second screenshot does allow the agents to add any additional notes but it gives them at least a good starting point so that um they can kind of capture the it captures the salum points that are in the incident that it's detected and uh gives them opportunity to um to use that during the reassignment flows okay um next slide one other area we also looked at is during the chat interactions so when you have um uh we originally introduced this in in the September uh store leas which we then further added in the November but the initial uh Focus around the summarization here was during um a chat conversation with from an agent and an enduser when the chat is completed it would summarize that chat now um a lot of times what we find is the experience in terms of creating an incident or any specific tasks often happens when you're still in the conversation with the end user you kind of determined oh I need to create an incident for this enduser let me go ahead and create that while I still have them on and then pass the information back so we've enhanced the chat summarization capabilities that we had originally launched in the November time frame or sorry we enhanced it in the November time frame adding on to what we l released in the September um store release and what we did here was allow the capability where on creation of the task the uh the current chat up to the that point in time will be summarized and added into the short description and description Fields um so as we kind of look at that um we we're we're continuing to look at where we can have more opportunities where this can be in in uh within the workflow the existing agents and we're going to continue to add more and more of these capabilities into our other itsm product areas such as problem MIM and uh change but this is just the the start of where we're building out these capabilities so I'll pause there for a second before um I jump over to some of the VA capabilities that we've launched in November see if there's any questions there is one Jason which is my client is concerned about the Privacy what about pii data of users when using nowes so it's a great question uh pii and all type of proprietary data for any customer is handled just the same way as we've always handled it in the platform so prior to let's say any of the training that we do with customers who have opted in for their data we strip out any pii or any sensitive information overall prior to sending it to llm for data sharing or training purposes we also have quite a bit of flexib availables for customers so they can actually not only in their Contracting have flexibility but after 30 days up to 30 days after their contract they can opt out of data sharing completely to be able to train the models overall Jason anything to add there no I think you covered it um I think most of uh most of any of the pi related um concerns they're they still kind of adhere to our previous AI technology so if you've used any of our things like task intelligence predictive intelligence in the past or even our virtual agent with nlu um all of those uh privacy and security requirements should be respected all right um let me jump into newer capability that we did launch in the November time frame um if you're familiar with our existing virtual agent experience it was an nlu powered um virtual agent um if you're not familiar with it we're introducing a newer virtual experience that's initially starting to be focused more on content so it's going to be focused on initially focused on content driven experience um around your knowledge base as well as your catalog items so kind of gives you the opportunity to take advantage of what you already have existing on your instance and making it a conversational experience and in the past um what we found is kind of setting up an NL based virtual agent does take quite a bit of time there's is um building out topics building out intent and utterances for those that have tried to set that up in the past so what we've really kind of done in the the new virtual agent experien is make that setup quick simple kind of laser focused on content initially we will have topic support in the subsequent releases but initially it's really just to help customers get up to speed very quickly um in the past it might have taken weeks or even months to get up and running whereas now you could have this up as soon as 30 30 minutes or an hour um to get the capabilities activated so let me kind of just start with showing um kind of the initial setup um in the terms of activating the capabilities um so just just for those that haven't seen this before um I'll start kind of looking at this page this is the now assist admin console and from here you can activate some of the uh skills around summarization that we showed a little bit earlier ear or talked to a little bit earlier um I'm going to drill down a little bit more specifically into the U virtual agent side for now um just uh just to kind of focus in on this conversation so you'll kind of notice within the itsm side you've got chat there's some knowledge uh management capabilities and then incident skills we're going to go look at the virtual agent one specifically in this case I've actually already activated it but if you were starting net new you would come in here look at the virtual agent section go into the details of it and from here there you'll notice that there's a couple of active skills that I've turned on um this would be in an off state if you're looking at a brand new instance but the main skills here are really focused around catalog so looking at your existing catalog and making them conversational and then also um looking at your knowledge based sources you can turn one on or the other or both on in this case I've turned both both of these on once you've activated these you kind of will go into this setup experience and this is a guided quick setup experience I've already gone through the setup but I'll walk you through a little bit of the steps just to show um what that experience looks like so you've got these activated skills you can kind of manage them and turn them on as needed um you will then determine which information sources you want to point to so in this case I've got a kind of the defaulted filter but you can go in and modify this as needed if you wanted to only show potentially like the It catalog or maybe an HR catalog um or you could look at all the catalog items um you can kind of set up your filters as needed here um same goes for the knowledge base if you have specific knowledge based articles that you want to surface or a segment of knowledge based articles that you want to filter on that are made available in the virtual agent they can be filtered here as well so so once you've kind of determined the sources of information you want those to point to you can do some light branding on the experience this is sort of the default um you can add some customized scening messages and some of the closing messages but if you kind of used our existing noou experience um this is sort of like a lightweight abstraction of that existing branding capability just to allow you some light changes now once you've kind of configured the brand it's a matter of determining what locations you want this virtual agent to be deployed um so right now it it's surfacing the different options most of them are portals you could launch it on the knowledge portal employee Center or service portal these are typically the ones that we're seeing customers surface these um so in this case I've got employee Center service portal and some other ones selected once I've kind of finished selecting that I can go and review it from here here I can look at you know just a summary of what I've selected I can also test the virtual agent in a panel it'll be a popup panel that I'll show there I'll go in and show the requestor experience after this but um but you'll kind of be able to see okay what have I selected for The Branding what information sources have I Chosen and then also where are the display locations um and then also there are some uh performance metrics that we can also look at some lightweight based um reporting analytics so once I've configured all of these this will be turning it on so in this case I've already configured this so this is this is showing turn off but this is um the quick setup of the admin experience for virtual agent that we've uh introduced for the November time frame um any questions before I move off of the admin experience as you can see it's pretty quick um it's meant to be a much more pair down version there's no longer the need to really create a lot of the topics ahead of time to be a to get this up and running I'm not seeing any questions come here Jason but I know one of the questions that has come in the past is uh what makes a a catalog item conversational yeah what makes it conversational is um typically it's looking at specific variable types that you're using within the catalog item um when you're thinking about a lot of the uh custom client scripts or custom UI policies that's usually when those don't don't uh lend well to a conversational experience but there are definitely opportunities to you know use different components that make it more conversational also if you've got you know something like 20 questions or something that probably doesn't lend as well to a conversational experience so those typically will either pop out as a inline modal depending on what experience you're looking at or it'll just launch a new page so it's still all supported in the virtual agent it's just the manner in which it's rendered to the end user um I do see a few other questions here does now assist uh Now assistant support multiple languages for the in user today this is only English language support is going to be coming uh next year I believe if I remember correctly there are about five languages that are coming live in u u late q1 if I remember correctly and then there's going to be much much more added as well that's right yep right now this is the first release of the virtual agent uh gen powered virtual agent so it is um currently only English good question okay let me um switch really quick over to just um the newer experience um I'm on the employees Service uh Center portal and um I can essentially engage in the chatbot here and um what I'll do is just walk through a couple of scenarios one will be more triggering the Q&A function so this will look at um I'm going to just ask something simple like how do I set up automatic replies and um what this will do is look and see in the knowledge base that's on service now to that I'm pointing to if there are articles that are related to this particular topic and if there are it will try to tease out if it finds a good match within that article on the steps so in this case you'll see that it's found this particular snippet within this article that's referenced down here on how to set up the automatic replies so this just kind of gives you an example how this content driven approach um can be pretty powerful in terms of not just pointing traditionally where you would look at a whole article pointing a user at a KB it tries to see if there's information within that KB that it thinks it has a high confidence and matching um and answering the particular question by the enduser um so that's just an example of the knowledge based portion let me go through an example of a catalog ordering experience um so I'm going to go in and order an iPhone 15 Pro with 256 gigs of storage and I need it by tomorrow so one of the things you'll see that's um unique now with our gen powered VA is um the ability to do slot filling so in the past even if you provided additional pieces of information that could be answered within the catalog it would um it wouldn't pre-fill that information so I'm going to go ahead and uh type this in and it's going to look for any catalog items that might fulfill my request and and uh it finds that I have an iPhone 15 catalog item that I can order I'm going to go ahead and get started and get started will basically kick into the catalog ordering experience if it's conversational it will ask the questions here if it's something that's using you know custom scripts or behind the scenes custom UI policies it'll render as um some sort of Link or potentially like a pop out so in this case um it's going through the questions that um are required as part of this catalog item I'm going to go ahead and just give it a business justification and then as part of this um I'm going to go ahead and say I don't need carrier service since I already have it and then um it's identified that I am the user here I can go ahead and change it if I need to but this is for myself so I will go ahead and proceive with this and then um I'm not going to exchange an old device and so now I will choose a location for the delivery and this is pulling from the valid locations that are associated with the offices for this and so what you'll see here is because now we're we're filling in some of the field ahead of time without it even prompting me to ask it provides a summary of the catalog ordering so in this case you notice that I didn't type in the Pro I didn't type in the color I wanted I didn't type in the storage um or at least it didn't ask me those I type that in right on the initial request um so it's filled those in because it's determined that those match specific pieces of the um uh of the catalog Kellog item and then and then also next thing you're looking at um in this case I'm going to go in and um create some changes maybe I don't want this to go to this location maybe I don't want it to to be uh I want a different color um I'm going to go ahead and make changes here and then from here I'm going to just type in something like hey actually I need the max amount of space and also a different color so let's go with the black Pro Max so in this case now what I'm going to do is it'll essentially pick up hey it understands that Max you know using large language models it understands what the max amount means so it's going to look for the storage type that has the highest amount in this case 512 that's available it's changed the color to Black it's also given me the prax model versus the original pro model and then from this um now I'm I'm happy with the changes that I've made and I'll go ahead and submit it so in the past providing multiple pieces of information and in our um older catalog experience was a bit of a challenge and so a lot of those experiences are a little bit more natural in conversation you can also ask some follow-up questions related to um the questions that you have assuming that that information is available in that catalog item so I've gone ahead and requested the item and you see now that this has completed the catac ordering experience um I'll pause there see if there's any questions any comments feel free to use the chat there are two questions that came up Jason I did answer one but I'll bring it up anyways because I think it's a good question is um could you go over the the part where uh we can replace the now assist powered virtual agent functionality uh to replace place the nlu based VA yeah so uh let me take a look is that question going over the replacement of it in terms of the portal so so one thing I'll I'll um I'll just answer is that today you can have both existing nlu um based VA and llm based VAs on the same instance the limitation there is that it can only surface on one particular portal so in this case for example if you have an existing no base one that's deployed on your service portal if you were to deploy the new llm based one you'd have to choose a different portal or it would have to replace the existing no based virtual agent on that portal so think of them almost as two different um chat bots on that instance and that there is a just coming from a concern from a privacy perspective does the llm use use instance specific data to train like catalog item details or KB article details uh there is uh typically with the the training um it's only it's not being trained like you would think of traditionally like an nou model it's being trained um uh centrally by um our teams and there is an opt-in opt out section in the now assist uh in the now assist setting virtual agents setting so if you come in here um you'll see there is a way to toggle this and you can determine if you want to allow that to be used as part of that Central training for the model for this data sharing piece yep thank you Jason one more question is any catalog item can be orderable through VA do we have any specific configuration that needs to be enabled to make them available so short answer is there are constraints at the moment which is that the catalog item needs to be con configur um conversational which means that you have any scripting or if find any UI elements that are associated with it that would not be able to use with the the current this virtual agent experience Jon anything to add there uh no but uh but yeah there is um there is a I kind of alluded to this earlier there's certain typ typ of variable types that are supported we've got a a comprehensive list of that documented as well and we are adding more guardrails and capabilities around surfacing what catalog items are supported and what aren't we do have pretty good coverage on most of the components that are pretty standard ones um and we're continuing to add more and more support for catalog items so the there are some probably changes and tweaks to make them conversational if you have ones that are commonly used but um we we have pretty we I think we cover Mo almost 60% of most of the catalog components today and some just are um are UI based driven ones that aren't going to lend well to conversation conversational experiences that I mentioned earlier okay covered all the questions so far team any or anybody else more questions we got 15 minutes left probably Jason you and I probably should have come prepared with a few maybe dad jokes as well so we could keep going and Alive okay one question did come up good so dad dad jokes always get people going can this now assist be integrated into s so good question so now assist is actually by default available in the service operations workspace and the UI 16 experience both ways so irrespective of which experience your customers or the customer is using now assyst is available yeah and just to maybe since we've got some time I've got another instance pulled up here um we showed more of the screenshots uh of it um but you can kind of see I'm in the service operations workspace in the August and September remember time frame we delivered a capability where as I kind of alluded to earlier it's more of a manual type summarization where an agent has to go in and do it what we what we've done is we've still we're still allowing the summarization to happen here but this is more of a kind of a quick um manual trigger by an agent and um a lot of times agents are potentially either looking at the details tab maybe they're not on the overview tab so we wanted to to take that same summarization experience and bring it in line of the experiences so in this case let's say I'm trying to um like reassign the incident or something to that effect it'll kick that off or in this case maybe I want to resolve this incident so if I click here for the resolution assuming that there's enough information on this to generate resolution notes um it will go ahead and fill out what it knows from the incident and from here I can then add additional information I will need to populate the code but you notice that there was a little bit of a an kind of a UI uh information that kind of popped up that showed that this was filled by analysis so the agent knows so those are the things that you know from a from analysis capability standpoint we are uh heavily heavily invested in the agent productivity side we see a lot of um interesting use cases around generation and summarization that can definitely help um augment the agent's experience in resolving incidents okay if there's um nothing else um shashan do you have any wrap-up slides we can cover I think for the next sessions do you want to kick there was actually one more question Jason yet came in so one I'll cover that one but just to finish what said also about the there is a 200 word limit threshold that we currently have set up so we believe that for us to be able to provide any type of meaningful summary in any of these scenarios you want to be able to have at least those number of words in the incident so when Jason was referring to limits that's what it is set up currently now you can go reduce it but then the op then there there's chances of hallucinations or or things of that nature as well so not recommended there's one more question Jason that also came which is will it be trained to answer natural language questions about the cmdb in the future yeah it's a great question David that is definitely something we're looking at in in general it's just to answer any questions that are available on data that's on the instances we'll be creating um additional skills past what we have today which is around you know incident summarization resolution node generation KB generation we'll have aition skills that are you know asking information about cmdb good question any last minute questions um we were going to wrap up otherwise we covered all the material that we had for today for the November release so I wanted to make sure we do a quick check with everyone still have 10 more minutes left so don't hesitate to ask us questions about what you saw today if you have questions even about the previous Vancouver release we're happy to answer them as well we'll give another minute or so to see what comes in he uh yeah we um we're always happy to help and if there are more questions that anybody has um I would encourage you to reach out to Jason Rai where we're at your service to be able to answer those questions uh but with that uh we'll wrap up for the session today uh we're just I'm going to take the the control from you thank you for the demo and thanks everyone for joining I know for some of you it's u a little bit late in the in the night or in the evening so we super appreciate you taking the time and coming chatting with us with now assist and we look forward to having you in U one of the future sessions as well