good afternoon thank you for joining us today kosol technology would like to welcome you to our service now Federal Tech talk from itsm to gen AI fueling Federal Innovation with service now loow code joining us today with service now we have Mark Maurice senior advisory solution consultant Joe Kim Senior solution sales executive and Steve live Federal Enterprise architect before we get started I would like to quickly go over a few housekeeping items please note that all lines have been muted to reduce any kind of background noise during the presentation if you have any questions throughout the webinar please feel free to use the Q&A feature on the bottom of your screen and we will do our best to answer by the end of the presentation or follow up with you offline following the webinar this webinar is being recorded and a copy will be emailed to you just to tell you a little bit about koft we are a trusted government IT solutions provider supporting public sector organizations across federal state and local government agencies including education and Healthcare markets at this time I'd like to hand the floor over to our speakers Mark the floor is all yours uh good afternoon everyone Mark here and um welcome to this webinar Lo code from itm to J on the service now platform so to get started we're going to start with our introductions and then we're going to go into the Foundation investment of Lo code of application development on the service now platform and we'll go into the details of our gener strategy and how is going to boost our platform's users productivity right so my name is Mark Moes I a senior advisory solution consultant on the Creator workflow team supporting all our federal customers and I'm joined today by Joe and I'm joined also by Steve Lee hi Steve Lee um service now Federal platform architect uh thank you guys and uh so before we get started we're going to talk about Innovations and Innovations often you know happen you know have some plans to be delivered uh so I just want to mention uh with the safe har not is that we're GNA make forward-looking statements and so as usual you know plans may change so before you make any decisions with service now uh reach out to your team and reach out to us we'll be we'll be uh more than happy to answer any question we have about the world map but make sure that you don't make any decisions based on what we show today right so just to get started and to reframe the discussion a little bit I think it's important to understand you know we work on a day-to-day basis with a lot of federal customers and many of them have been very successful on the itm platform but they they think that you know many of them can only think that you know service now does only itsm but service a platform was actually built by for founder here with a vision to allow anyone to build any kind of meaningful workflow applications so the foundation is in the product to be to allow anyone to build applications fast and so as we started to build on on that Vision we added more and more tools that we go and invested on those tools from the guided app creator that allowed very easily to create an application with mobile studio and templates to get started very fast on building a mobile applications on the platform and more components and integration tools directly on the service now platform so Joe yeah so sorry about that so as I was saying since 2015 uh the state department has implemented hundreds of mission critical custom apps using App engine uh to meet the ongoing and evolving needs of State Department employees these apps include the department wide my services and my data Custom Service portals uh and these apps provide access to hundreds of different state department Enterprise service requests like the example shown here on the slide the state department has started to and continues to build on service now AB engine because of the ease of building workflows and approval Logic the faster speed of development our governance capabilities to help them track app portfolio usage uh as well as having a portal framework to deliver new services and a mobile application for a better user experience so as service now is grown we continued to build acquire and integ technology all with the same purpose to automate work better faster and smarter we've expanded development Studio to app engine Studio to allow more people with all different skill sets to build on the platform and soon after we introduced app engine management center for app and user governance which is a big differentiator for low code platforms and so as you can see from all those uh Investments already uh what we have in mind with the platform is to allow anyone to build applications on the platform but also making sure that as you go and step to all the the stages of building an application to the platform you have governance every step of the way so from the intake where we provide an intake form to make sure that the right application is being built that there is no spoil in the agency and then as you go and build that you have the tools to be able to validate the correctness of that application and that the right Integrations are secure and you have the proper Access Control to to make sure that the data is as private as you need to be and then as you go and you have all the tools to make sure that you can oversee the overall deployment and management of those applications and some application you may not need so you have the tool to understand the usage of those application do you need to continue maintaining those applications or do you need to re hire them and so those tools have allowed Us by the simplic it and power to host many many workshops and we are here today in the service net office with a hosting another two-day workshop with many federal agencies where we where people get a hands-on experience with the tool they're able to ideate on not only existing to uh Solutions but also think about all the uh different ideas that they have and Implement their own application on the service now platform so want to extend an invitation to everyone here on the webinar today to reach out to us register for the next webinar which is on December 13 and come experience the power of the platform right and one of those customer that came to one of the workshop is the chapain services and Joe's going to talk a little bit about this this amazing story yep that's correct Mark yeah you know keep in mind you know chaplains are not technical by Nature know their principal purpose is to promote spiritual and personal wellbeing of service members the U service now Crea workflows team helped the chaplain staff with planning designing and quickly building a mobile friendly app to help chaplain and sailors manage their spiritual journey know this is an excellent example of how anyone even non-technical folks can build on the platform you know even Mission critical apps using a drag and drop uh Dev development tools and governance features yeah and I think that Jo you mentioned also about that story that people in the chap and services don't have a technical one yeah absolutely but you know to that point mark you know through the workshop they were able to quickly realize how much easier it is to build custom apps using App engine instead of other local platforms and the workshop really helped them realize ways to ideate understand how service now can help solve Agency problems you know they were able to successfully evaluate and outof the-box flexible and modern building experience without the need for coding and the workshop really gave them an opportunity to execute on their low code strategy and citizen development framework know none of this would have been possible without attending that workshop and the app's outcome is actually helping promote better ways to support a positive emotional and spiritual work environment so that our service members can continue to serve and protect our nation so we really encourage everyone on the line here today to reach out to us about how we can organize a workshop to help benefit your agency's loow coost strategy Um this can be at your office or even one of the service now locations yeah and you know this is amazing what happens when you put people together in the same room and we go into whiteboards and we talk about different ideas about how we can improve this process and what also what the pain points uh really that people are experiencing in those processes right it's so much more efficient then looking at you know I'm sure that everybody on the line here loves dealing with request for proposals and requirements and so on but if you put people in the same room um things get done much faster so I again encourage everyone to learn about this and come to a workshop so this is all the Innovation that we have built so far that we are building but you know right now we've also invested a lot into the Next Generation and the next of generation and like everybody probably has heard right you you've used chat GPT you've used generative AI to generate some type of content we've been looking for years about how we could apply this to accelerate the application development cycle and Gardner says that by 2025 80% of app developments will use some kind of gen cogeneration and that's just one piece of the the puzal you're going to see that you know gner is looking at just one the C generation part but you're going to see here um I'm going to turn it over to see that's going to talk who's going to talk about service Nows AI strategy and all the capabilities that we are that we've been working on and then we going introduce right Steve to you thank you H thanks uh Mark so this is a perfect slide okay so and that's a key part right service now right has seen the whole AI um market and environment progressed over time we saw gen AI coming a number of years ago so six years ago we made a intentional commitment right to Innovation and to strategically either acquire or build uh top AI Talent you can see here based on this timeline we spent a significant amount of capital right and um internal resources building what you're about to see which is our geni capabilities and add it on top of it are already existing high level or sorry uh wide portfolio of AI capabilities so we really invested our time and made sure that we're here and we're on the Forefront of this geni technology next one so at service now our main strategy right and again service not is not new to it and we are in a great position and our main three strategies here area of focus is service personalization right knowledge synthesis and content generation right service personalization be able to go to a service portal and interact using natural language and ask questions and get a personalized response because our AI knows um the user what they're entitled to what they have access to and that knowledge uh synthesis right being able to recognize there's a knowledge Gap right within your organization and be able to use our combination of machine learning AI search to generate knowledge articles and fill that Gap and then content generation and that's what you typically see out in the market with that chat GTP like um experience right being able to summarize um content from record or from knowledge articles for catalog requests so that's a main and then here's a poll question we were GNA send out to everybody so if everyone take a look at So currently it looks like everyone is uh a majority of the people are we have no plans we're just considering J geni at this time and and you know to be honest as I've been speaking with a number of different customers within the federal Marketplace that's the exact same thing they're looking they're you know they're kind of falling into the same category where they're hearing a lot of uh news and talk about gen so you know they're looking at it as a opportunity to get ready for what's coming up down the pipe so they're doing a lot of the research now so you know it's glad to see that that still holds true here within this particular webinar yeah it's pretty exciting to see also you know 40% of people plan to in C gen in the upcoming year so we'll be ready for you guys so close the the poll on to the next slide all right so our three main areas to focus how are we going to achieve that right and and this is how we're we're going to acheve service now alterative ai strategy is centered around providing Broad and secure large learning models and you know as some of you know or if you don't know behind that chat GPT is a what we call a large learning uh sorry large language model right which is learning and um taking in your response and providing a result back so service now we've are we're actually um going to provide two capabilities the ability to use a generative AI controller to integrate with the general purpose um large learning models out there sorry large language models already out there um you might have heard of some which is that open API or the Azure open API um but we also will be um building different spokes Integrations with other partners that have llms which are Google Cloud and then we have what service now are calling domain specific llms and we're naming them specifically now llms right and those domain llms are going to be specifically trained on our industry and customer data right so that way it generates the most relevant and accurate answers based on the service now's capabilities so you know in in kind of going over this customers have two options if they're building bespoke application or custom applications they can use our generative AI controller to do the integration with thirdparty vendor llm such as Azure open I and all of our other box um generative AI capabilities will be relying and working and using our now LLS any questions there no so and I will stress with the slide again the Safe Harbor slide so this as of now because again generative AI space is moving so fast a notational architecture of how it would look for our federal customers um as you know you can see across the top there right customers within our GCC NC data centers we are going to replicate our now llms right our which is our domain specific llms are into that secure U regulated data center right at the high level you can see here if you use any of our outbox now assist skills as we're calling our geni capabilities now assist skills it will actually send the request to our now llms within that same Data Center and then if for example a transaction gets too high or we're having too many requests come in we do have or we plan on having a cloud burst capabilities and that is again will be hosted in the secure data center it actually will be using Azure um secure government data centers service now will still own and control that um now llm it's just hosted in a um Azure data center any of the conversational gen cap that customer wants to use out of the box will you be using our OEM llm and that is essentially a copy right of azure's open AI um chat GTP model just the model itself not the data just the model we will be using our own uh data to actually train that model for conversations and then you can see here on the left in the red right if customers you choose to use our controller to integrate with thirdparty ones they have the same capabilities and right now as it is as it stands um we do have access a control go to market for self-hosted customers but again they have to purchase their own Hardware specific hardware and host those large learning models uh yeah large language models um in their own infrastructure and again just to reite just to um all this is notational It's subject to change all right so this is like I said the the the Gen environment is moving fast new capabilities are coming out right and new deploy and this is example here how we are developing um gen capabilities of across the platform and across each of the workflows right and you can see here um in this first release we just did with Vancouver right we have for example case summarization and that goes toward that whole content generation and that same uh case or incident summarization Works across all workflows and then specifically if we just focus on the creative workflows which is you know a key part of this Workshop today right text to code being able to help custo developers generate code using best practices using the same code that the r LM was trained on using our actual developers internal product teams development the code that they write we used it to train our llms any questions so let let's talk about that content generation here right and that's kind of one of the things we talk about and that's what you most likely see out when people use that chat GTP like user experience so we have that here on our same portal where you use natural language to ask a question it uses our combination of our AI search to go out and look for the most relevant material within your instance and then send that information to our now llm which then generates a response and you can see here from the screenshot that top section where it's powered by now assist that's the response from R now llm right and the key portion here where it's going to help a lot is when you do General search you have a lot of knowledge articles you have a lot of catalog you have to waste clicks going into trust to find what you need the summarization is going to be the key changer and reducing that click time and getting the right information to the uh to you at the right time so the case summarization and this is going to be a big part of our agent to actual um agent to agent collaboration or even agent to um end user customer um reduction for me right I recently had a user experience where I was dealing with a you know internet service provider where I was talking chatting with the virtual agent and then you know that workflow ended up transferring into a live agent the live agent said hey I need a couple minutes to review you know the this case so that way we can proceed and I can help you well with our case summarization and a big part of it is when that actual transfer going to happen we will generate a summary of that whole conversation right as well as the information within that case and present it to that agent so that way they don't have to burn that time saying hey let give me a chance to review this before we go it will be at the fingertips for them they will have that summarization and the same likewise for agent to agent transfer when you're transferring to another agents they don't have to spend time sifting through the case through different tabs through the different um Fields just to see what you did right they will have that summarization right there for them yeah this can save so much time for users on the platform yep and the conversational so this is the key the key part that same analysis search will be available within the actual virtual agent um the other big thing that is coming to that virtual agent is the ability for that quick turnaround when you do a quick search and it recommends a catalog item for example you know typically what now customers have to do will is click on that catalog item that's recommended by the virtual agent it reroutes them takes them to the actual catalog item itself with the New Gen features utilizing our natural language structures right that virtual agent will be able to ask you the questions that are related to that actual catalog item so that way you don't have to be rerouted and go to that catalog to fill it yourself that virtual agent will automatically ask you in natural conversation hey provide me your name okay what group are you and all that information will be prompted to you and it will become a natural conversation with the virtual agent just to complete that actual catalog item so you heard me talk about earlier the Gen generative AI controller right and this is going to be key for um using and our Creator capabilities right being able to build custom applications using the controller to generate or build custom um virtual agent topics or even integrate our actual existing or um gener AI cap into the actual applications and again using Integrations to Azure open AI or Google they will be able to bring those same gen AI caps to your custom applications so thank you Steve um so we talk a lot about uh some of the Comm that coming across the platform so what I want to go into a little bit more is how the capabilities that are going to make creators even more productive with Gen and what we do that we look at the different personas because when when we talk about creators right there are so many people that interact on building Innovations in the platform you may have developers administrators someone that may be more specialized into maybe a process analyst or someone that's more specialized into building Integrations for other in the platform to consume so if we look at this developer I would gen there are some studies that say that you can actually reduce app Dev development Time by 50% or more right and so we look at other other use cases for an admin how we can help them manage the deployment and the governance of applications also looking at from a process analyst or Integrations how we can drive and help them the the build and the definition of those processes right that's all we're going to go into so this is right now current as of Vancouver release we have now available inside your Vancouver release flow recommendation and we just used that into our Workshop today in the service office um where as you go and building a flow the flow is aware of what you're trying to build is aware of how other flows of similar natur have been built and is recommending for you the next best action to use in your workflow right so it's learning and it's helping you as you go and text to code right so there are some areas in the in your application where let's say most of it is going to be no code very easy to configure but there might be some areas where you need to write a couple of script and we know that you know you need to have a certain knowledge today to be able to write those script but with Gen you can quickly write a comment describing what you want the J to write for you and hit enter and it will generate the script to be able for like here to qu interet older than two years right so there is no need to spend a lot of time on the documentation or learning more about coding here it's quite the scripts for you gives you a jump start right it's always a good idea to make your prompt for J more very very specific and understand that it's a it's a it's a it's a a jump start foure script it's not it might not be the the end result but it's gonna increase your productivity tenfold right and text process it's actually released today on November 16 where I'll be able to describe what my process is about let's say it could be a financial disclosure process it could be you know a uh security access control request it could be anything uh that you may think about and is going to generate inside the service now workflow environment all the different stages and steps and activities that need to take place or are usually uh performed inside this workflow like here we we prompted the text to process about an application about for transaction risk evaluation and it buil all the all the steps so I will encourage everyone to look into your Vancouver and reach out to your account team and talk about how you can turn this on into your instance so we talk about flow recommendation text to code customerization and text to process I love for everybody here to give us a little bit of uh what features in generative AI you're looking forward to using and you can use any number of them um usually the the more the better uh but uh I would love to get you input those are available in in Vancouver as we talk we just build a workflow with the uh the attendees of our workshop and the recommendations were popping and we started talking about it that that was that started to spin off some very interesting discussion right even if you if you don't have a plan for using J in the very next future this is a very interesting discussion to have about where it could be applied in your agency we talked about right now those four key features of J available in Vancouver and we're continuing to innovate on the platform uh we so some of that was already released in September for Utah patch two uh and store two and then we adding additional capabilities from accelerating the application development to augmenting the user experience what can we do some of the features that Steve talked about how we can provide a better user interaction with the platform for the end user and also how can we accelerate automation so we talking about automation assist with the flow but how can we make this even better and expanding the Integrations and gen with allowing you to extract more data from uh any kind of documents that you have you want to process because we know that in federal we have a lot of documents that we still need to store to some extent but there are ways that we can help with the service down platform to extract a vast amount of data and accelerate any kind of manual process that is required today to process those document right so in the future uh and this is what we're working on today so what you see are uh descriptions mockup um some are way more advanced than here uh but the in the future so don't make any ping decision about those feature yet uh but this is where the platform is going it's it's going to help you bring more um guidance into the tour uh what's amazing is that we started testing some of those features for example when I'm building an integration the integration script here is aware of what integration I'm working with so if I'm working with a Microsoft teams integration it knows the variables that I need to use so I don't need to wander inside my integration or go to the documentation about that team spoke integration for example directly from there the gener is aware of what service n has built and is going to make it easier for every Creator in the platform to build much faster directly inside their integration so next another type of user in the plat platform someone that may just need a dashboard and just need a very a quick way to add a new data visualization right so we already augmented today with the workspace builder in appen studio the ability for people to build their own dashboard very quickly but if you need something even faster is is how can we bring this to the next level is adding a dashboard talking to a gen agent and having them understand what we want to build and having them to build help us build this directly in the dev environment right and again we have built-in governance all the way from managing build the build of the application to managing the deployment of this app so when we do this kind of automated build this is within the context of the government that is put in place so if I'm asking someone to build something for me this is my Dev instance and the app engine management center is going to help me make make sure that this app is built properly because you never know why the the Gen is an amazing tool to help you get a jump start in your development and build faster but you want to double check everything that's being built on your behalf right so again we help people build faster with built-in governance that we have today with aage M Management Center we bringing more governance to automate the work of admins how they can be more aware what's being build or how can we make their work uh faster to manage the deployment or request from users about the application that need to be deployed to from da staging and production and so what we're bringing next is really the next level of innovation is all the tools from App Eng Studio to app engine management center is how we can bring building and apps to the future with nalysis for everyone where with just a question or a description of the problem we can have people like here describing how do they want to make a form and having the now assist tool that is going to bring them into appen studio is going to promp them for the data that they need what type of information or picture and why commend that most of the time when such application is requested there is a a need for a headshot and the the AI is understanding the request but not just doing what the user is asking is also providing guidance into uh all the past application that we built very similar and is helping build the UI of this application it's helping build the data model so that everyone directly on the service n platform with Creator workflow has the ability to build the first application in minutes all right so we are bringing this uh in the next future for you to be able to build this application very quickly so last pulling question we have whole questions with fixed um questions I'd love to hear from you um you've heard about generi even if you don't have a plan to use it in the very next future but are there any other use cases you're thinking about in your agency that you might be wondering if service now can help you with right and so we would like to have your uh feedback so we can get back to you and give you what you need to to continue innovating on the service now platform before we move into the Q&A i' love I have those QA codes here that allow you to very quickly go to the sites for registering for Innovation Workshop get a personal developer instance a personal developer instance is something that service now does that very unique um not lot of vendors do that it allows you to get your own personal service now instance uh just for you to be able to learn to be able to um try things on your own and and learn more as you go and those instances don't expire if you keep using them and then ified uh as a service now developer go onto or now learning website and uh start learning uh many of those trainings are free Steve you have something to say yeah I just want to note um the personal developer instances the geni capabilities will not work with those um so if that's what you're then you know contact UM Mark or the account team directly if they're interested in seeing and you you know having a a test run of the Gen capabilities yeah it's a great point because you know there was a lot of Investments that are behind gen there was a lot of resources that are needed so we can't really make that happen on on the PDI for or thousand and thousand of instances but if you reach out to us we we can make this happen for you um great so now um we can take any question are there any question I can see there are Zoom questions so at this time we'd like to open the floor for questions please use the Q&A pod on the bottom of your screen to ask any questions that you might have it looks like there's a question mark how are weights leveraged in the model does an agency or customer have the ability to apply their own values against the model from their internal smmes so you want to take that one yeah so that the key thing here is um with the federal customers and being in a regulated market they will their data will not be used to train any of our models at all they they will there'll be they'll be by default um opted out of of um sending data to our models for training does that answer your question hopefully yes so I the question was can that data be used so you know we might need to do a followup if people want to use their own data with service now large language models um we can inser in more details another question for example how can I ensure that a user who does not have access to a piece of information cannot use gen I to go around their access limitations and get the piece of information that's a great question I think St I think it's built in the in the architecture right yeah so the geni will comply with the existing ACLS and the security on a platform so if they don't have access to that particular record or knowledge article um gen will not summarize it yeah it's a great point because you know when he talks about the power of the service a platform everything is on one architecture if you set up Access Control parameters in your instance everything in the platform is going to respect those if you define um data anony anonymization Wars in your platform everything in your platform is going is going to respect those parameters another question he asked where is that written down are you asking if there is a official document about our J use of customers data or how the the security parameters in service now prevent people using Ser gen to circumvent that the lad uhhuh yeah I don't think we have an official document but it's just built in inside the service now platform um feel free to reach out to us we we can give you additional details yeah I'm gonna I'll look into and check to make sure because all the information should be published on our docs now in terms of um how it's built yeah and we have also a gen page on service now.com there's an official FAQ that we can share later on with the audience next question we have another question everything demonstrated today is available in the Vancouver release so everything that was marked Vancouver is available in Vancouver there were a couple of things that were marked as future that service know is working on and just because it's Vancouver patch too someone asked is the release with generative AI available on the federal cloud yet at currently not right now we are still waiting for the hardware to um to be received before we can actually uh Implement that any more questions on the line um I need to specifically see the IAL vendor documentation that confirms that the acl's and other security restrictions are applied in Vancouver and Beyond yeah we can get that to you so uh we follow up yeah security is par amounts on the SEC on the service now platform so uh if we don't have the documents already um we'll have it next question looks like we have another question will service now support other large language models besides open Ai and Aur yeah there was in the architecture slide that Steve presented uh we are working with Google cloud model and also the the way that the integration to those models are built um you know service now is open in a way that we provide those equations out of the box but if you if you have your own let's say large language model or if you if there is a partner that service now is not connecting today with then you can build your own integration as well someone asked can we post this recording internally I think that recording is going to be available publicly so you can always redirect people to this recording another question do we have to use the Gen AI controller to use gen AI in service now so like everything in service now you don't necessarily have to use it um you can you can always make your own service now is a great platform to create your own Integrations your own components and it's built for that to allow you to bring your own ideas through the platform but there was always a decision that you need to make at some point where there's going to be a cost Associated to creating your own versus the cost of using service now's own already made pre-built integration with those models right those Integrations are done through the generative AI model and it gives you flexibility where if you want to change from one large language model to another you don't have to change your code you just need to change a parameter inside the service now gen controller so there there was going to be a lot of POS and counts for you to evaluate before you know making a decision but short answer you can always build your own inside service now the answer is that sometimes it's not because you can that you should so um talk to us and we can help you make those decisions we have a question is there a timeline as to when the Gen AI will be available in GC PC I just heard there's Hardware that still needs to be received and installed correct uh at this time we can communicate any fixed timeline but we're trying to do that as soon as possible we know this is important for at least I number was 40 50% of the people here on that phone so we know it's important we're working hard on it we have another question how does the recent om memo and developing policies impact the product road map um I mean the ref to the executive order or just is there a specific one from om executive order yeah gotcha um yes so the key part of the executive order right is um that we're you know there's multiple parts that we're working on but the the the main part of kind of the open AI being transparent and and uh available for kind of public review um service now has already gone down that path right within our docs environment right we do Post what we call model cards and those model cards will give you a breakdown of you know how our models work the benefits the data how it's used um the risks the limitations all the information is will be posted based on our models and if you go to our doc site and look up the our analysis model cards you'll see the links to those publicly poish sorry publicly P published model cards um so that's just the one key aspect I know that we were we're working with on the to get that done for the as part of that executive um order yeah we can send the link also afterwards to the audience any more question I love them uh please have them coming up no more questions on the line I'm not seeing any more questions okay thank you very much um do you want to close the the call yes thank you team I'd like to thank all of our participants as well and our speakers for being with us today we hope the information that you received during this webinar has been helpful to you if you have any further questions or would like to request more information please feel free to reach out and thank you again and have a great day thank you everyone one