welcome everyone to today's process optimization Academy session focused on automation Discovery now this is a topic we've touched on briefly in other Academy sessions but really haven't done a deep dive on and and more recently we've had a few conversations where someone would say hey there looks to be a ton of opportunity here in this automation Discovery tab I'm just not really sure what this part of it means or where to start or what I should do next with the information that I'm seeing here so we thought it was time to get some content out there to help folks better understand the value that that portion of process optimization or the process optimization workbench was was providing them so really I just want to thank everyone who's provided the feedback keep that feedback coming we want to make sure that these these sessions are providing value so if there's a topic you want to cover please let us know what whether it's uh via email or a post on the community or a comment on the process optimization landing page I want to make sure that these are are valuable for you now before we get to automation Discovery we we have to get the academy housekeeping out of the way for our newcomers and those who have stumbled upon one of our recordings for the first time but after this one I'm sure you'll go back and watch some of the others the purpose of these sessions are to help us all get better educated on servicenows in platform process mining capabilities uh we'll provide how to's practical guidance and use cases product updates and just really provide a place where we can exchange ideas and collaborate so we all continue to get better how the sessions work if this is your first time we have a scheduled topic that we covered usually for the first 30 40 minutes or so then we open it up for questions on the chosen topic for that day and then you open it up for just more general questions above and beyond the day's topic sort of in an office hours type feel uh these sessions are recorded and eventually posted on the community so if you you don't want to be recorded in any way maybe just put your questions in the chat rather than coming off from you and then finally feel free to post your questions in that chat section or the Q a section and we'll do our best to answer them and then for those situations where we don't have an answer for your question immediately we'll follow up maybe with the post on the process optimization optimization product Hub on the community or use them as a topic for a future session kind of like this one this is the slide that says anything we say and do here today cannot be held against us in the court of law it's also the slide that says hey is part of these conversations if we make any forward-looking statements about what may be coming in future releases of the platform you should take them as such and make no decisions based on them because those things are always subject to change I'm sure you've read the fine print really important because I know that Ranjit is going to be talking a little bit about our roadmap when it um comes to automation Discovery and finding automation opportunities on the platform so I am lucky today to have a special guest one of the leaders behind our in-platform process mining solution and someone who's been involved in bringing a number of different solutions on servicenow platform to life manjit Singh manjit I know you've joined us before but uh do you mind doing it just a quick intro for everyone yeah absolutely uh hi everyone my name is manjeet Singh I live in Santa Clara California with my beautiful wife and two kids um I've been with company for about seven years now in product management and I love two things the most building out great products and staying perfectly fit mentally and and physically so that's the goal so we have at least one thing in common I'll let the audience figure out which one all right uh so what we're gonna do here today as always we'll do a little bit of process mining 101 then uh manjit's gonna walk us through the why and what of automation Discovery uh he'll do some demos talk a little bit about use cases that uh you know in terms of how to use the information to identify automation opportunities on the platform they'll give us some information about where we're headed uh with automation Discovery and new automation opportunities we'll get to your questions then we'll wrap up with hey you loved what you saw how do you get started here's some resources for you and we'll talk a little bit about future sessions to come so as always the quick process mining 101 for our newcomers with everything we do there's a designed and desired path in our minds for for how it should play out and how it should work like when we started thinking about this session here today I totally thought that menji would be here at least two minutes ahead of time but didn't play out that way but I had a vision for how it was going to play out and when we designed things we designed for both efficiency as well as completeness to provide the best experience possible for anyone and everyone involved unfortunately what we design isn't always what ends up happening in reality the reality is that not all the work flowing through the processes on the servicenow platform is going to take the optimal path and that's going to have a negative impact on both the experience that the employee may be requesting the service is having as well as the people that are actually doing the work behind the scenes now identifying what's actually happening within our workflows and then improving them isn't always the easiest thing to do now what in platform process mining or our product named process optimization allows us to do is use the audit log data that is generated as records moved through a given workflow on the servicenow platform and it allows us to then use that data to create a visual representation of what is actually happening within a given process and this new level of visibility helps us accelerate our ability to identify process and efficiencies non-conformant process activities and in general Improvement opportunities ways from which we can get better process mining gives us the ability to answer process questions that have historically been very challenging to answer so whereas like traditional analytics will allow us to answer a lot of the what questions about our processes process mining helps us answer a lot more of the whys so things like where is our process stuck where is the rework happening where are tickets ping-ponging between groups and where aren't we conforming to the process that we've designed at the very beginning of this whole thing right these are all things that have historically been very difficult to answer and then more importantly act upon um and now they're available to us in in a couple of clicks which you've seen in in other sessions we here at servicenow like to say process optimization is going to get you to the why behind all of those kpis that you have on your dashboards today now in past sessions we've talked a lot about how our in-platform approach helps organizations move from just doing one-off process mining science projects to being able to include process mining as part of their continual or process Improvement routines or Cadence that they have inside of the organization um and this is because the platform provides connected solutions that help at all four phases of a process Improvement workflow if you will or a program so at the detect phase the analyze phase the improve and monitor phase now today's topic is bucketed on this slide in the analyze section but as you'll see as manjit goes through this it actually touches on all four phases in one way or another and with that I'm gonna I'm gonna pass it over to you manjit to kind of just take us through automation Discovery it's power and kind of some of the use cases that are out there all right so let me share my screen and get uh get going here so Dan can you confirm is my PowerPoint up here uh it is awesome all right so I'm calling this section continuous automation Discovery uh this is uh very cool this topic is very close to my heart more recently because uh as a company we have a dedicated strategic Focus around this term called hyper automation because the platform has now platform has evolved so much and we have so many Lego boxes already in the platform when we think about automation different kinds of Automation and as they are coming together as we are building out a unified experience how do you use them in a complete life cycle which includes discovery which means hey let's find out where do we have the automation where automation can or should be applied and then using the right tool set or technique to guide the user further hey this is a an opportunity probably you can uh deflect it through self-service or a service portal or maybe this opportunity there is a good candidate for RPA and then able to do this in a in an ongoing basis um that's the goal we have and provide all the tools capabilities and Coe around it on a single platform that's the the goal we are moving forward with Will Foods full speed so let me uh make it more concrete because this session is about process optimization so what role in platform process mining plays and like where in this circle so I think the most important part here is the Discovery section uh towards the bottom here so this is where um I will walk you and show you guys that what we have done is using in platform process Mining and machine learning and AI based automation Discovery they're coming together to help you discover automation opportunities across all the workflows and processes that run on servicenow platform for example itsm CSM HR and all these different processes that power up these workflow um when we think about some of the outcomes um there's some kpis that I'm hearing again and again uh in different cios uh or CEO um conversation things like hey we want to increase our self-service rate or uh we already have itsm Pro which provides us a bunch of automation capab Tools in in the platform why our automation rate is so low how can we increase our process automation rate from let's say 10 to 20 so those kind of goals and targets I'm seeing more oftenly being talked about um and then you know the kpis like hey can we quantify the Savings in terms of the the cost or the time now the question becomes is uh how to make this a reality and another way to look at it is uh what already exists in the platform okay um so I'm taking this one example of case management and if we see in the life cycle of a case when the case gets created what are some of the capabilities that can help more specifically answer on the self-service angle many times the goal is maybe avoid the need to create the case completely you know so serve your user using different techniques like virtual agents your knowledge base uh better AI search that can help find people you know common questions and queries that I have for example today morning I searched it's like do we have the Wednesday holiday you know because Thanksgiving week and uh VA responded me yes so things like I think some of those common otherwise you know those might become a case that somebody has to work on the second part of the journey is let's say the case got created okay uh and then the then the goal becomes is uh how can we help your agents itsm agent or customer service agent to help resolve those issues as fast as possible or how can we find the right agent with the right skill set so that you know when the classification happen or categorization happen it gets routed to the the correct team or or the uh the uh the person who will Who can resolve it fast because all these leads to the faster resolutions better SLA and higher customer satisfaction and when I think about let's say the service owner or uh you know that Persona um they want to have the full visibility when from the monitoring standpoint so that's where um you know our performance or disk capability and as well as the process optimizing capability comes and how to provide visibility into what's what's happening behind the workflow where the bottlenecks inefficiency or conformance issues and then again you go back to this continuous Improvement uh uh full closed loop so as you can see on this slide there's these different boxes so these are the capabilities uh most of them are automation specific capability then the question becomes is how can we find this automation uh potential automation opportunity we analyze them we bet them and work on the on the opportunity that that have the highest Roi uh quickly another uh way to look at kind of step-by-step process before I go into you know the demo part another so um if you look at the step one and step two so that's where I'll Focus the most today take any process that runs on servicenow do the assessment on the potential automation opportunities uh there are a few different techniques we will look at and as you find those automation you with one click you send them to the your automation Coe or continual Improvement management and then rest of it is obvious you know any life cycle that you see you know somebody has to make those kind of process changes or and then you want to monitor for your Roi uh zooming in a bit specifically on the process optimization part because as I mentioned in this hyper automation goal process optimization really dominates in the discovery phase uh of the uh of the automation discovery so the simplest view is uh you take any servicenow workflow and then Within 5 to 15 minutes you can map it out quickly so you will you will be looking at your process map you will be looking at uh all the different insights uh and then as you do the analysis that I'll go into detail uh in a in a bit uh you will find those potential automation candidate and then those candidate could be uh is it best fit for some will be for virtual agents or there will be for uh our AI ml capabilities like we have a similarity model we have clustering models and classification models we also have RPA and dock Intel capability in the platform so the the goals that we are moving towards is uh we are doing a good job in certain areas like detecting VA and other opportunities so uh we want to detect opportunities for all the tool sets okay um so if we look at the automation Discovery feature that is now embedded within process optimization there are two different techniques that we are using at a higher level one is a process mining based technique so this is I would say one of the differentiator we have we have built a finding engine uh and if you feed some of the uh the patterns around like hey these are the common touch points and um inefficiencies that happen in a process um and then those finding rules for discover uh those areas as as your finding insights card so that's the approach number one couple of example if you quickly read through it hey I'm interested in looking at the what are the how many times my ticket always go from like L your assignment group level one to level two and if it is happening with a very high volume and if the complexity is low can I detect this Behavior because that is a potential automation opportunity to automate that step the second Technique we have is uh is a machine learning based technique so the way it works is uh um we have specific intent that comes with this capability and uh it looks through let's say uh if we see the incident process example all the text data in this sort description and description and trying to find out what are those incidents and what is user asking or complaining about and then see if those types of issue can be automated using using the tools that we have so let's drill down a little bit more like to make it more concrete so uh what the process optimization native capability will help you do and this machine learning technique embedded within the same workbench so two quick example first one is uh this manual trousing we have seen a lot of customers like okay we have this sizable uh theme and the only thing they do is pretty much you know incident comes into a shared mailbox or somewhere and then they uh they you know those those teams analyze those and you know a few minutes and then send it to the right team uh and I think this is a good example because we have a very mature classification model which can Now understand the all the incoming incidents and help Auto classify so if in this case in this example uh we run it through our internal use case if you see the incidents were coming into it how that's where my mouse is and it was going up to this group but once we have the classification model uh it's applied onto that particular step uh the auto classification is happening uh using this machine learning model so and we have achieved kind of uh you know significant success where people are pretty happy it's like hey 80 of the classification is correct now and some of those savings can be can be Quantified in terms of the time saved and the cost saved um another example we can look at it is uh uh this is a very common uh issue that we are seeing uh anytime Dan and I are on a call with customer and it's like hey you are running the PO first time we want to see and one issue we always see is uh waiting for customer information or waiting for user information um and many times this field get misused in a way you know teams knowledge issues skills issue the root cause could be you know uh different but this is a another example an opportunity where uh is if once after you do the drill down is like potentially using maybe an RPA and then how you can nurse the people how you can you know inform that hey this thing is sitting in your queue uh or uh you know the uh analyzing some of the work notes uh and guiding the user okay so with that you know the context of the like how we are thinking about the continuous automation discovery the two techniques I talked about um let's look at the individual features uh that will help us discover the automation opportunities when when you start using process optimization so first is we will look at the finding rule based Insight uh I think this is as I said this is one of the key differentiator we have in the product once configured right this works beautifully and we are providing a lot of you know the good finding rules as part of the different content pack itsm CSM and HR the second feature that process optimization has a is a variation analysis feature so we'll see how this feature is helpful in analyzing the potential automation opportunity of course if you look at the ml based Discovery I mentioned and then some of the new enhancement that we are bringing in is uh is look for specific patterns at the process level and here are some good examples for you to uh to think about when you start let's say analyzing your process for automation discovery that is always a good idea for to look for these patterns okay for example high volume because you don't want to apply automation if you're starting let's say uh your journey on just dealing with one-off anomalies and others I think those comes later as as your process gets more standardized and steam line the first guidance we provide is you look for high volume low complexity uh low standard deviations the less rework or reputations uh and scenarios because a lot of times people are getting involved so most likely you know it is a uh it is a complex uh problem or uh approvals you can include um many times like load durations are a good good bucket to look at first uh uh and I think look at the areas where sometime before work the the your life cycle is going to work in progress and specifically thinking about can you help self-solve the users even if let's say the incident or case is created Channel analysis is a great one we are uh seeing um a lot of inside customers have discovered while doing Channel analysis looking at how many how these tickets are coming through email and phone and portal and what's the volume what's the the average duration what's the uh the standard duration so I think this is pretty easy to do in the uh in the process optimizing product today uh user journey is where we are also focusing on um so today we show you the process map of what uh kind of your task Journey based on the audit history and now we are also thinking about extending into going Beyond like you know at the user interface level like as people comes to service portal in the future releases we will be stitching together that user's Journey as well so that you can see hey these many people came into service portal these many people did the search they ended up in this KB article or 10 of them got diverted to Virtual agents uh I think that's uh coming together with your task level process Journey will be very comprehensive analysis because then you you can look at the journey before those incidents and tasks are getting created and think about how you can uh provide more self-service and increase automation okay all right so I guess lots of slideware let me switch into the demo section and I'll show you uh um more particularly let's if I go back few slide we will look into at least these four pieces if we have time I'll have some more examples later in the slides that we can go uh let me take a pause if you have any com questions comments then um let's hear from the group if if anything uh was not clear or if I may need to clarify here from the chat uh just somebody was asking about whether the slides would be available and of course we take the slides and the recording and they get posted to the community I put a link in the chat to where you'll be able to find this stuff later but nothing nothing um about the solution as of yet but feel free to post those questions or come off mute and ask them okay sounds good yeah so um so let's let's look at the uh the first piece there um we in one of the previous session we covered these finding rules how to set it up uh and then uh for this session um I looked at few use cases and examples uh and uh uh My Demo data is limited but I was able to create some of these finding cards and as you mind the process you can see if these matching conditions are there then uh these finding cards become very handy even for people who don't have too much of the process knowledge or or like the data science skills because these finding card highlights what is the opportunity what is the potential impact in terms of the time in terms of the volume right in my slide I was showing some of the patterns to look for right in terms of the volume the complexity the uh the number of people involved so the way for example uh if we look at one of the how these finding cards are coming to life is uh um what you do is you go to the process configuration okay and then this will bring up a process specific so in this case we are looking at incident uh as a process and down if you scroll down the bottom section is is this where you define your finding rules uh through these finding definitions um so for example uh a files for more details on how to kind of set up these findings we have another session I would say you you refer to that but let me show you a couple of examples that how I have set it up uh I'll open up this one let's say the manual trousing example that I I had in my slide so in this case you can sequence that is like I I'm looking for a behavior as soon as a incident or case gets created drag that in the journey and if it was assigned to L1 and then eventually followed by to L2 and you can also Define the time durations like within three minutes five minutes because typically L1 is the analysis is quick quickly you know if it's sitting in their queue for multiple days a different issue so there is a Time duration constraint as well I haven't put it yet but just to give you an idea that you create um these like sequence one sequence two um save those and next time when you will run the mining all those matching opportunities will show up as an inside card here um and and this one as we uh we are going to spend a lot of because um as uh hundreds of we have I think 300 customers that are using process optimization today there's a lot of learning coming in around like the common inefficiencies and one bucket in those common inefficiencies would be automation opportunities uh and so the future content pack will have more and more examples like this that we are uh we are creating a library where then then those will be part of the uh the newer content pack um so now the question becomes is okay this this is you know maybe a potential opportunity what actions I can take uh so then what you do is you start analyzing these opportunities further right so you click on the card it takes you to the map um then you start to look at the uh the variation analysis uh that you have uh I mean in my example I have very limited data but this the I think the one key aspect of the variation analysis is look for what is the volume in a given process path uh like for example this one this is just one record let me go to my another instance I think this will give you the better so I'll go to the same spot let's say you're looking at a finding card that gives you hey there's a potential visual agent opportunity for example and then you come to the variation analysis and then what you do is uh using the filter shorted by most record okay so as you can see there's 11 2.4 K records are uh only going through these two step process if I pick this for example uh this is a step whatever the reason is it directly got created and directly moved into resolved and closed um this then what you do is okay maybe if this is not interesting let me let me include the second one so this is the path that it's taking from here assigned to resolve so maybe this one is a little interesting I'll you know click on it see more data apply transitions um so I would say is uh the the thought process where uh we are using internally and like discovering the automation opportunity is is combination of your finding cards and then also looking at the variation analysis because this gives you an idea around the volume the time durations and then the complexity or like a how many steps uh that particular set of Records and uh in one customer scenario we found out like there are 40 of the volume was taking a very straightforward path uh a very high volume like you know the 300 000 uh customer cases um and then they found that particular opportunity very interesting and then um I think they they apply two techniques one is advanced work assignment to because the work was not getting in the queue and then some ba opportunity okay um the another uh if you think that okay it is finding cards are you know uh you can set these up in here good news is we are making this capability much and much easier in fact the part of the finding we are exposing it even outside of the uh outside of the finding card so let me quickly show you uh how this functionality has evolved so we have a concept of transitions when you go in the uh go in the process map and what we did is we bring in some of these finding as part of the transition so let's take a scenario I'm a process owner start analyzing this my process 3.8k record but then I want to look for I'm going back to my slide I'll pick one or two kind of uh let's say I'm I want to look at some of these patterns and maybe with the combination of who how many teams are touching these cases so then what you can do is with this transition logic you don't have to even create a finding card you can create your query right here um and uh and this makes it much much easier so let me show you one example for example let's say I say I'm looking well I'm interested in looking at incident that moved to let's say uh that went all the way to resolve are closed uh so let's take that example incident state is resolved and maybe first occurrence then what you say is I'll add next one uh incident state it came back to let's say work in progress okay uh in progress and all incident and the good thing is you can Define some of these conditions that were previously part of the finding rules followed by eventually followed by or not followed by so now you can think of your own scenario for example um I'm interested in I have these multiple themes like we have team in San Diego Santa Clara uh in our India idic office and some of the cases that we analyzed is hey because teams these teams work kind of around the clock and then they wanted to see how the work is getting passed around between those teams and then can it be you know there is some automation opportunity so uh in that case probably you'll not do with State you can do with the uh the uh other like if you have assignment group or as part of the activity uh another great thing is you can Define your constraint right here uh I think time based constraint is I'm interested in if an incident comes back from a result to onekin progress within two hours uh that may be an interesting one because we said it's resolved but then the user said no it's not solved so the good thing is this within the map without even going through the finding those you can think about the these potential you know hypothetical hypothetical scenario that you have and create that condition right here uh and uh you know it's a concept of chain so think of these conditions are coming together and you'll say apply those conditions uh you know with this button okay so I think a great technique to when you think about these scenarios you can Define like anytime you see a time that's where the constraint field that we have uh between you know one day or three days or or number of reputations you can Define like anything that is two days or three day um so uh coming back to the variation analysis one more useful uh analysis to I think that that should be done is I as I mentioned you this what what you are seeing here is all your routes okay how many if I clear let's say that my all the routes and if I sh if we as you have seen if I show you the entire picture of the process uh this spaghetti mat right because this one is very hard to analyze and that's where you come to the variation analysis and we have a filter here okay so then you can say is I'm interested in the key analyzing my cases for automation opportunity if my case is is greater than let's say uh 200 because I care about volume because you know more volume and if my automation if this is an automation opportunity the better Roi then you can say is uh this and I'm interest looking for the complexity so one way that one field that gives you the indication of complexity is steps so then what you do is if my step is uh let's say less than five okay so the good thing is uh you can filter because these number of routes if you see where my mouse is it's more than three thousand and it's like okay there's two it's too much but I want to really go into this anything that high in volume low in complexity and maybe probably the time Factor as well average duration is more than one week as you apply this then what you will have is a filtered list as you can see out of those three thousand now this is my first target analysis so I only have to analyze this 16 variations because and the good thing is the 16 comprises of more than 50 percent of the volume right so you want to go for the areas that that are kind of the quick win for you when you think about the automation analysis okay let me pause for questions I ventil back and forth so just want to make sure uh everybody was able to follow along and if you have questions please ask nothing specifically around automation Discovery a question around uh centers of excellence and stuff but I think we should hold off on that to make sure we get to the automation Discovery piece uh correct yeah um and I've posted in the the chat we did do a session on kind of planning a process mining program and talking a little bit about scaling process mining at an organization so I did post in the chat uh that Academy session for just in case we don't get to it today but I think uh just want to make sure we get to the automation Discovery stuff before we dive into other topics today sounds good okay so the what I covered so far is few different ways that are there like the approach one right you look at the finding rule you look at the variation analysis you can further drill down uh this uh you know the build onto records or you can do the uh the text-based clustering now next let's look at because this is the most interesting this is where uh uh we we have tested this on on lots of customer data and customer have run it by themselves as a standalone and now the good thing is uh this discovery technique is part of the uh the PO World trench so the way it works is uh once you have the uh automation opportunity enabled and let me show you how easy is it to turn it on so when you go to the process optimization configuration all the plug-in and dependency everything is is automatically enabled for activated for you so the only thing that that is needed uh is when you go to process configuration for a given process what we are looking here is incident go to the automation Discovery Tab and then this is the check box so in case if you don't see the tab in when you go in a workbench come here and check this off um check this basically or mark this on there is very default configuration like it looks for the the text data on the short description if if you want you know for these common processes like incident problem change so description works really well but if you have some other let's say the app engine type scenario or maybe a different field uh you go ahead and analyze it here um and then uh good thing is if I check this or to run this uh automation Discovery then I don't even have to run it manually so by default this is kind of auto run a quick note on this taxonomy because on right now we have the taxonomy for itsm the HR taxonomy is coming in in Utah release HR and actually icon taxonomy some of those AI op scenarios and then the team is working on bringing the CSM taxonomy so think of taxonomy is a list of pre-built intent um that when the discovery happens uh let me actually go to the the actual UI so you can you can see what I'm talking about so going back to the workbench uh like on for this process I already have the automation Discovery checked when I come onto this tab this is the kind of the output you will see so on the top you have these three cards that tells you potential number of automation opportunity uh if you click on that I icon it tells you what it means uh the second is the number of deflection because this like that if you see we we are looking at 22 000 records so what is telling me is uh 10 000 of those can be possibly deflected and then the last card is the estimated saving for the top 10 candidates okay so since on my configuration I set it auto run uh then I don't have to do anything I'll mine my process the project model and then this tab will be automatically populated for me so now as part of the journey uh you know some analysis as I said I'll do it here and then the uh I'll come into the automation opportunity and my question is what the based on the text Data analysis of these 22 000 records what are the opportunities that the the ml based Discovery has has detected for me so if I scroll it down uh you will see these two tabs first tab is automation opportunity and then second is not categorized so let me explain this so automation this bucket is based on the if uh if you have the taxonomy available or not for example in itsm on this instance I have the taxonomy so try to match all those intents with these uh these different clusters or or the bucket of incident so for example in this bucket I have one uh 1800 plus records uh some more stats how many what is the mttr assignment group uh what what are the you know who is working on that and look at this field here virtual agent ready so what it is telling me is that this opportunity is not only that it's a VA candidate but I already have the VA content pack installed so which means you know the the this is the uh the good opportunity to uh to automate through VA in some cases where I clicked on it uh we track the actions um right so for example when I was looking at this one uh I can see that this the VA topic it is linked to is unlock password account because if you see the the what the the cool thing about this this feature is it analyzes all your uh your same process data that you're looking in the map and then it looks for this opportunity uh you uh based on the the text so it's telling me it's related to the user account unlock and then here's a cool thing I I already have a VA topic available for for us too uh to implement and then uh when it comes down to the implementation you can go to the uh I think to my this my build is not latest so it's one one more card you will see here is the uh open the topic recommendations and that's where you can go and and start uh it takes you to the the actual VA and then the topic where you can attach these records and they will start automating so then probably will we can link the the video in the uh that there is a dedicated video on the uh the journey let's say the topic is there how to attach a new topic or then um for because the few scenarios that we are telling virtual agent ready and if it's not a VA ready uh then how do you you know uh link it by yourself so those capabilities are there as part of our kind of nlu workbench now um so now what you do is uh uh the one is okay you know the one important feature I should call out here is uh because this is integrated into process optimization now one question you may be asking is hey can I see the process Behind these uh right because I'm I'm in my process workbench can I see how these 1800 records are solved today from end to end in a process map yes so that can be done you see this link here open process map if I click onto this so what happens is uh it takes that ml uh that bucket those 1800 records and then it filtered out my process for only those records I'm now in this view I'm not looking at my 22 22 000 records what I'm looking at is those 1800 records um and now I can see just as a you know the second part of the analysis hey is there any more complexity involved um how much times time do we have between let's say if I put the duration metric and this is getting a little more confidence is whether this opportunity is worth uh you know spending time on further or making a proposal that we should automate um so that's I think another benefit of looking at what the the clustering and the automation Discovery technique is telling you and then also looking at the the process behind it side by side right into one uh one what bench UI um now coming back to the uh let's say if you don't have the taxonomy available for your table or kind of your process area um in that case what will happen is you might you will not see anything under the first bucket uh the fallback technique that we have is uh is a clustering uh without the without the intent taxonomy so which is pretty useful I mean this makes it a little bit more you know targeted it tells you something is virtual agent ready and then uh in the future releases we will look for is it a RPA ready and dock Intel ready and other other techniques that we have but I would say this one is still pretty useful uh because it will not give you the uh those you know the whether you have a virtual agent uh uh intent or not but it can still tell you that uh uh what other clusters that you have and what what is what is the actually the you know it's for example if I look at these uh Records this this is these are about telepresence and then the question becomes hey can should we create a topic for telepresence or and maybe drill down into these records or look at the process Behind These so I think it then it comes down to the analysis the good thing is uh all this is coming together you don't have to uh you know do it in a standalone and then look at the uh the process behind it all the variation analysis and then uh and then of course the power of finding cards I would say just going to chime in real quick man you know that that scenario where you have like just the clustering um and we might not have the matches on inside of automation Discovery a common common use case uh is just to mine the the tickets that are coming in categorized as other right um we we all get uh we all get a bunch of those I'm guilty of it here at servicenow I am that lazy person that doesn't look at categorizing it when I'm filling out the form I just say other um and then put something in but that those buckets of others the clustering portion of the solution there uh really helps you get an understanding of kind of hey are there any volume based opportunities in those other areas that I can that I can start acting upon um and improving the service a little bit so that's that's a good use case for the the other scenario for those lazy folks like me that are out there yeah that's a good one then yeah um um so in in I mean in the remaining part of the slide deck and we will we will share this uh we have put together a few more ideas and then things to look for um so for example uh ping pong we mentioned uh uh it's always interesting to look at your P3 and P4 cases because usually the we found a lot of opportunities there uh in in our case management because people were really focusing on P1 and P2 uh but there's this large volume of these low priority tickets uh but those are important from the uh your employee experience and and that standpoint um uh and as so we will uh be extending uh this automation Discovery to RPM dock Intel and the way we are approaching is for RPA it's the same technique you know we look for the the specific intent and it's like hey if I have a password reset RPA bot available and If password reset is happening in a third-party system like OCTA or Citrix or you know your Mainframe then we can look at this those set of you know those requests or or incidents and then automate using RPA documental is interesting one uh we are seeing a lot of opportunity in not all the workflows but I think the the workflows that are very attachment slash documents heavy uh case management we are saying good use cases there um we have some new workflows around procurement and accounts payable so that's where we see this doc Intel Discovery automation Discovery will be uh will be very useful for uh for the customers and the way we are looking at is okay uh what type of documents are used and can those documents be handled through our doc intro and the OCR technology and if yes then those opportunities will be surfaced up um yeah so this is what the basically the slide for uh the automation Discovery tab that we saw this another example of you know the uh this is actually a slide that came from the the P3 and P4 scenarios uh it's a before and after State how this was happening it before but then we implemented uh VA uh how fast this has become from 10 minutes to 30 second um so this this is an idea to how to kind of think about when you look at your the good thing is process mining tells you these you know the steps now because which was earlier very hard to look at how many steps are involved when you will have to ask your your users or your customers but now you don't have to ask you can say we have seven steps today and can we bring it down to Two Steps From kind of the manual touch standpoint um thinking about some of the topics the VA topics and then the uh the taxonomy that we have so these two slides the good thing is uh this library is growing uh with every release we have 80 plus different pre-built VA specific automation um and which makes the uh you know uh this whole ml based automation Discovery more and more useful um all right so let me just uh then try to kind of wrap it up here uh so the last piece maybe I can show is um what is the kind of another Discovery okay you find automation opportunity at the process level and then you submit them in the uh as part of your uh your this experience where uh you know Sim initiative and if you have the automation Co installed then you will see one more action which I have in this slide automation opportunity so you will um it's it's it basically you have a choice now whether you want to track them as same opportunity or if you want to use the um another purpose-built uh Coe location we call it an automation Center but the goal there is uh whatever wherever you are tracking I mean try to make it as a as a a central place because uh one of the goals would be not to do it in silos and not to build this RPA or VA in like in in a different teams it has to be you know the uh more Central so you can increase reusability you can increase your knowledge sharing and then otherwise I see these automation effort they they happen I mean the teams Kickstart but then there's uh you know a lot of effort on getting wasted and uh um and then the good thing is uh that automation Center has some pre-built dashboard and kpi so that continue to show you the automation achieved through uh RPA and those as you apply them into the process um a quick look at the roadmap um so we as I said we have it for uh today in even like before Tokyo automation Discovery for virtual agents the itom is um the item discovery came in Tokyo and then in Utah release we are working on adding those taxonomy and then kind of the the rest of the pieces for CSM uh and uh and the goal is to uh deliver hyper automation on one single platform then back to you I think we only have one or two minute left that's quite all right thank you Manji and you know a couple of questions came in during um during your portion here uh that uh were things that we may have covered in in other sessions so I've posted um links in the chat to those other sessions in which we covered something specifically let's say like that automation Center piece that you just uh you just discussed there right um real quick uh just want to open it up are there any questions that anyone has that they want to ask in the time that we have left if not I'm just going to quickly wrap things up uh for those that obviously understand some of you might have to to drop at the top of the hour but for the recording I want to get some of the logistical things in here uh for those of you that this was your first exposure to process optimization and now you're saying to yourself man not only do I love process optimization I would love to get the benefit of that automation Discovery capability and start understanding my data a little bit better and where there are opportunities to to improve and automate as part of the workflow what you do next is you go to now learning and you take the process optimization Essentials course it's about 90 minutes if you put it on two times speeds you get through it in 45 minutes just saying um and then the next thing that you're going to do is you're going to activate the plugins in your instance so you'll go in and you'll activate or in your sub prod instance you'll activate the process optimization plugin and then depending on the solution that you're using whether it be itsm HR customer service management it'll activate the associated process optimization content pack that content pack by the way remember the configuration that manjeet showed and what you had app for automation Discovery and checking the Box those configurations those records They come as part of that content pack so it's super important that you enable the appropriate content pack for your workflows there's tons of other resources that are out there I really hope that you visit the product Hub process optimization product Hub on the community form that'll get you links or if you search up servicenow process optimization FAQ on Google it'll get you to a page a landing page which has a ton of information and links to all of these other additional resources so search servicenow process optimization FAQ you'll get a page that's got all these links available to you last chance for romance and for questions anybody got any uh questions that they want to throw out there already that it was too too easy or it was no we just had a lot of content a lot a lot of content there um and the beautiful thing is that we now have it recorded so people can go back and watch it um at a high speed or a slow speed depending on what they want to do now I just want to wish everyone that's watching this we are coming up on the Thanksgiving holiday everyone that celebrates have a fantastic Thanksgiving and um next week when we reconvene after the Thanksgiving holiday uh right now the plan is to cover process optimization and the request process um and then go into interpreting interconnected processes and then the follow-up session after the new year will be on slas but we are always looking for ideas um and as you've noticed we've adjusted the content based on people asking hey we'd love to learn more about this topic so if you do have ideas for a topic that you want to cover please post them in the community or reach out to me directly I'd love to hear from you um that way we get the content out there that people are interested in seeing and learning about rather than us just making up what we think is important to you so it's all about you make sure you get us some ideas for the content other great academies out there that if you're not haven't checked them out already please go check these out I'm in with that thank you for your time and attendance manjit thank you for for stepping up and providing us a lot of information about automation Discovery and how to find automation opportunities really appreciate you helping us out here today thanks everyone thank you everyone see you next time