welcome to service now techbites bringing you an inside look at our products I'm your host Steve Miller today we're going to talk with product manager Mark Hopper about how service now is adding the power of gen to user experiences across the entire spectrum of our products stay tuned hi everyone today I'm happy to welcome Mark Hopper to our virtual Studio hey Mark thanks happy to be here how are you doing today doing great before we start talking about gen I wonder if you could tell our listeners just what it is you do here at service now I'm part of the platform organization my team covers outbound product management for the AI and generative AI engines within the platform okay do you work at a particular service now location or are you remote I work mostly remote I'm in the Bay Area east of Oakland love it over here lots of opportunities to be outside and mostly just here you know working for service now and spending time with my family and our Cody all right to set the stage for our chat I wonder if you could just outline for our listeners what service now overall approach is to AI it's a very popular topic with our customers and I think one thing that a lot of people don't realize is they're not new Concepts to service now so we've had AI Frameworks in our platform and in use within our applications since 2017 it's really uh a core element within the platform leveraging all the data that exists in the service now instances empowering like I said many workflows for a number of years generative AI also isn't new to service now we've been working on this for a few years through an acquisition of el and AI back in 2021 I think one thing that also comes up quite a bit as our strategy around AI is really not to provide them necessarily as just open Frameworks we really focus on the utility of these Technologies to deliver business outcomes with our customers that goes from identifying task Trends to automating repetitive tasks and also allowing end users to better self- serve themselves wherever they are specific to generative Ai and our initial strategy we really been focused on a few key pillars and those are supercharging agent productivity reimagining self- service and then a lot of exciting opportunities to democratize creation of workflows and coding so developing without code would be the third okay looking across our product landscape then in light of what you were just saying what would you say the top application for Gen aai is right now you know across those three pillars I think each customers they might prioritize them a little bit differently as part of this generative AI initiative we already have in market and in product capabilities that serve all three of those pillars that I mentioned but one of course like supercharging agent productivity we've seen a lot of customers deploying a whole Suite of skills that are powered by generative Ai and when I use the term agent it's really for people maybe that don't know service now that well so this could be like a field service technician or a level one or level two agent that would be sitting or covering your it help desk or your HR cases customer service operations management you get the idea anyway these agents they're responsible for delivering a resolution to an enduser issue or to a work order and what's cool about generative Ai and you said what are some of the use cases that people are deploying these fulfillers they they have a lot to consume to come up to speed they've got case history or chat history they may have to like search through manuals or knowledge based articles if one of the cases is handed over and you know as user you're probably familiar with big BRS where you call somebody they can't help you and they've got to transfer you then that second level agent has to come up to speed on everything so there's a whole set of generative AI functionality that can really enable these agents to come up the speed so much faster through summarizing all these different types of materials whether it's chats or manuals and search and whatnot and to give them guidance on how to resolve them on the other side of that equation when the job is finished the generative AI can automate the generation of closure notes which also improves like the overall system intelligence generative AI can take on the burden of generating future kind of like knowledge article that can help avoid similar issues so at the end of the day it really decreases resolution it saves costs and it just delivers a much better experience to both agents and end users so it really contributes a lot throughout the execution of the task on the ramp up stage and on actually executing the task itself and then at the end wrapping it up absolutely that particular application that sounds like a a real double wind for the company in the sense that it makes things a lot easier for the agents themselves and it cuts the cost of the operation for the company and it contributes all that knowledge and experience to the company for future tasks similar tasks going forward covers the whole range there yeah the business case driver around generative AI on that agent side is really strong so it's it's not uncommon that we'll have customers or new customers that are still operating by taking in cases through email or phone and so the cost benefit around transforming experience with generative AI is really powerful you also mentioned self-service what are we doing there in that self-service pillar we've shipped a number of skills that really reshape what that end user chatbot or search experience looks like it's just much richer much more intelligent in the past customers that are building out chatbot type experiences they have to script out and be much more specific about how that conversation will will flow it's what we would call a topic and you might think about it more as a use case and as part of that you'll have to script out every question and answer you'll have to train the model to understand all the different synonyms and different ways people could say it so it could be a lot of work but with generative AI the need to script out that conversation greatly diminishes and you also don't have to handle training in various language model in the case of say like a catalog request and that's catalog is a is probably one of the primary ways that people try to selfs serve the user wouldn't have to walk through this kind of like step-by-step guided conversation they can really just State out what their need is or their question and the generative AI engine can parse that entire statement and only follow up with questions on any missing Fields as an example a common use case is an end user in a company needing access to a particular software or maybe ordering a device so if a user ordering an iPhone a previous conversation might ask individual questions of each field that is required in the that catalog item so what color what size phone where to ship it what the business justification things like that and they might even require the person to go in a particular order with generative AI a user could just say I need to order a green largest iPhone with the biggest memory I can get get they can really put it in any of their own words and service now virtual agent now already know which phone model which memory size which color that you needed and so it's reply might be only that missing item which could be the business justification because we also already know where you live or where you work where what your default address is so the setup is also much easier as I mentioned in the past customers would have to go and build out conversations they would have to uh train models and whatnot since that's really already done with the conversational catalog you can really set up the virtual agent in as low as you know certainly under an hour some customers do it in their test environments in 15 minutes and they can start self-service immediately all right we've talked about agents and self-service at the top you also mentioned developers how are we using gen to help those folks yeah this is a a very exciting area because one of the things I often hear from customers is one of their consistent challenges is the ability to hire nowal proficient developers and they're kind of slowed in their ability to transform their business by creating custombuilt workflows and things like that customers want to do more with the platform and they want to expand the number of workflows they're running expand the number of the Departments that are leveraging the platform and so on you mentioned now assist so one of the things that we have is this product or Suite it's called now assist Creator we use the word creator for reason because it's really democratizing the way more people can build and create on service now your classic developer has to write a lot of JavaScript code and we have this skill or we call code generation and there a developer can express the needs of what they need to build in an application in a sentence or in a paragraph and now assist will generate service now code that can then be reviewed and deployed so for an experienced developer it greatly accelerates the ability for them to write code for a less experienced developer it starts to open up the ability create new applications and that entire Suite will not just generate code even higher level Frameworks right so we talked about catalog earlier the cool counterpart to that conversational catalog is a feature we have called catalog generation so now you know you can just describe in a sentence or a paragraph the need for some new catalog item and it'll generate the catalog item for you same thing for playbooks and workflows so it's a really powerful tool set and we're seeing a lot of traction with customers around that particular capability this is really great to hear because I know that reducing the amount of code expertise that would be required that's been a priority for the company for a long time it sounds like gen has just really turbocharged that effort absolutely all right we hear a lot about commercial gen tools like chat GPT things like that and those are making a big splash but those public tools aren't always so well suited for business applications how are we addressing that yeah we've heard from many customers that they're essentially uncomfortable with the concept of taking data off their service now instance and sending it to a third party public generative AI service so our strategy really has two important constructs and the first I know I've used this term a few times already is what we call skills so those are essentially out of the box ready to deploy use cases you can think about them as pre-built prompt that we've done extensive testing around how to like actually execute on behalf of that agent or in user plus some user interface and these are focused on very specific Enterprise oriented tasks again going back to even the earlier statement we made about how we focus around AI which is to deliver uh particular business outcomes the second construct to the skills is ENT price specific models um that are tuned and managed by service now running in service now's dedicated data centers and infrastructure so you would hear the term we talked about now assis being the umbrella brand for generative AI we also have now llm which is really our highle branding for all the various fine-tuned large language model service now is operating and these two constructs are closely coupled the skills paired with the model deliver those business outcomes that I was talking about earlier all right well when we were chatting earlier you mentioned that there was a lot of interest in gen at knowledge 24 among our partners can you tell us about that yeah it was front and center AI was a big focus of knowledge 24 we had record attendance many of our customers rely on our partners to deploy and manage their service now implementations so in addition to knowledge where we had Keynotes and a number of sessions with Partners around now assist we also held a set of dedicated webinars and boot camps with our partners that's really set records for attendance and you can actually already see many of our partners already working with now assist all right well that's great to to hear about the adoption side of things it sounds like our efforts in implementing ji across our Spectrum are paying off there I can imagine our listeners saying all right this all sounds great how do I get started what would be your advice there the customer looking to get started first I definitely recommend they connect with their account team but if they wanted to start exploring on their own we have a lot of resources that are published there's a series for example of now assist courses and certifications on our learning platform called now learning there's an entire generative AI section on our community site under platform AI forum and then we have other materials on our website and product docs Etc so there there's quite a bit out there for them to get their hands on well during our chat here here you've hinted a lot of different developments here at service now wonder if you could tell us a little more about that what else new is going on so one of the things that we're doing with generative AI That's a little different is we're iterating quite quickly and so we're releasing new skills and models every quarter and you'll continue to see that as a core part of our strategy if you're asking about what's coming in some of the other future developments if you attended knowledge 24 then they were treated to a lot of previews of you know a number of new innovations that are coming in 2024 and 2025 if you weren't there I'd encourage anyone that wants to see these to watch the keynote they're all posted actually on our website all the Keynotes highlight a new area of innovation you know across all these different pillars that I mentioned and some many exciting experiences whether it's multimodality autogeneration of content on behalf of the various users whole set of new skills in that kind of Creator category where you can write a text to create anything so whether that's analytics or applications there's just a lot coming there okay it sounds like the the pace is matching pace of geni in general it's something new every week yeah no doubt all right is there anything else you want to mention we're a platform company and we really encourage you know Innovation both with our customers and our partners it's important to note that generi it is an extensible system at service now so if there's other models that customers would like to connect to their instance that is an option it doesn't just have to be one of the now llms that service down manage so if there's particular model customers are running on Azure or Google or they have their own complete fine tun model you have the ability to augment now assist with that we also have quite a bit of tools that allow customers and partners to build and create their own custom skills in addition to all those kind of now assist skills that I talked about that are out of the box well before we wrap up I wonder if you could uh tell us where we can go to get more information on all this absolutely I mentioned earlier so on our website there's a a whole generative AI category that would be one area product documentation and the service now Community the other thing I would maybe a selfish promotion is my team runs what we call AI Academy which you can sign up for it on the community site but the replays are posted on YouTube and based on like high customer-based feedback of what the use cases that we see customers trying to do we actually build them out live on instances is we take questions and those replays on YouTube we often hear are a way that many people learn how to work with our products I'd also encourage people to follow AI Academy and we'd love to see you on the next one all right Mark I want to thank you very much for coming and joining us today uh this has been a great way to bring us all up to date on a topic that's moving forward very fast I really appreciate it thanks Steve to learn more about any service now topic take a look at our product documentation or knowledge base or our service now support YouTube channel Channel thanks for listening