well welcome everyone to today's process mining Academy session focused on multi-dimensional process Maps so multi-dimensional process Maps what they do is they they give us endtoend visibility and allow us to identify processing efficiencies and improvement improvement opportunities across multiple tables so think of a map linking incident problem and change or maybe requested items and their Associated tasks so this capability has been available since the Tokyo release and has definitely been a pretty highly requested Academy topic that I'm happy we're finally getting to cover but uh before we get to the main course let's quickly get the logistics out of the way for those that are new uh the purpose of these sessions are to help us all get better educated on service Nows in platform process mining capabilities we'll provide how-tos practical guidance and use cases product updates and really it's just just about exchanging ideas as a group and collaborating so we all continue to get better now how the sessions work we typically have a scheduled topic that we'll cover usually for the first 30 minutes I'm pretty sure today we'll go a little bit longer because I'm going to be configuring a couple of things uh then we'll open it up for questions on the chosen topic for that session then we open it up for questions above and beyond the day's topic sort of like an office hours um and and just so you know these sessions will be recorded and they event get posted on the community this will be I think our 21st recording that we'll be posting so no shortage of opportunities to learn a little bit more after today if you want um feel free to post your questions in the Q&A uh section I'll do my best during the session to answer them there's a couple of times today where we'll be hitting the mine button so that'll give us a few minutes maybe to to actually catch up on some of the questions and then for situations where we don't have an answer for your question immediately we usually follow up and post the answer either either to the recorded page on the community or the the page that's hosting the recording on the community or just to the community in general or maybe we use them as topics for future sessions now this is the slide that says hey anything I say and do here today can't be held against me in the court of law it's also the slide that says as part of these conversations if we happen to make any forward-looking statements about what may be coming in future releases you should take them as such make no decisions based on them because those things are always subject to change and I'm assuming you've all read the finding print uh for those that I've not met before my name is Dan Grady I'm part of the product team here so it's now focused on our process mining solution what we'll do here today we'll do that quick 101 for our our newcomers uh then we'll dive into the topic of multi-dimensional mining um or multi-dimensional map process Maps we'll run through and slide some common examples we'll talk a little bit about interpreting the map that's one of the the um more interesting things there's certain things that change when you start looking at multi-dimensional process Maps we'll look at some of the things that change about how you interact with the workbench to ask questions of these M multi-dimensional Maps then we'll jump into the instance and I'll just show you a couple of things that we talk about in the slides live we'll configure a couple of projects so we have that in the recording you see what goes into creating uh a multi-dimensional map we'll get your questions answered and then we'll wrap up with some of the additional resources available to you above and beyond today as always the quick 101 for the newcomers so with everything we do there's a designed and desired path in our minds for how it should work and how things should play out whether that be uh planning an event like these Academy sessions or a business process or workflow on the service now platform and when you design things you design for both completeness and efficiency to provide the best experience possible for as many people as possible unfortunately what we design isn't always what ends up happening in reality the reality is not all work is going to be flowing through the process and taking the optimal path and that's going to have a negative impact on the experience that both people requesting service are having as well as the people that are trying to deliver that service now identifying what is actually happening within our workflows and business processes and then improving them isn't always easy what process mining allows us to do is use the audit log data that is generated as records move through a given workflow 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 inefficiencies non-conformant process activities and ultimately Improvement opportunities process mining starts to give us the ability to answer process questions that have historically been pretty hard to answer so where traditional analytics will allow us to answer a lot of the what questions about our processes um process mining helps us answer the why so things like where is the process getting stuck where is there some sort of unnecessary rework happening where is work ping ponging between different groups or where aren't we conforming to what we've designed these are all things that have historically been a little bit of a challenge for people to figure out and then ultimately act upon and now as uh you've seen in past sessions and you'll see here today the answers to some of these questions are just a couple of clicks away uh we always like to say that process mining is going to get you to the why behind some of those higher level kpis and in doing so it empowers people to make datadriven process decisions or process Improvement decisions um and we can apply process mining to a single table like this incident example on the screen here um or multi-table processes which is the focus topic today so for example the request that has both items uh in one table on the service now platform and tasks uh this is a a very common use case for for um multi-dimensional process Maps right where you've got you want to identify the approval bottleneck at the request to requested item level and then the work related inefficiencies at let's say the task level this multi-dimensional approach is also powerful for HR life cycle events uh like onboarding so where you have that high level onboarding case and then a bunch of subtasks and requests involved in the process and we'll dig deeper into these today and just for those that have not uh attended prior sessions we can now also apply process mining to process data that's been imported into the service now platform from external applications like this example of a hiring process driven by data that was pulled from Smart recruiters um so just trying to get that end to-end visibility for our customers all righty let's dive deeper into those multi-dimensional process Maps so many workflows are just a series of interconnected processes and in order to understand where the inefficiencies in those processes lie it's sometimes useful to have endtoend visibility so we can not only see the bottlenecks within the individual processes themselves but also the potential and efficiencies as different processes start to interact with one another multi-dimensional process Maps allow you to stit together in multiple processes or tables through parent child or uh reference relationships so we can accomplish this type of analysis right the incoming and outgoing arcs between the processes are able to give us or help us get a sense of the incoming and outcoming volumes and velocities as work moves through the different subprocesses SE on within in a higher level workflow now there's some common examples where we' use multi-dimensional maps we already kind of talked about a couple of them but the first is with the request process where we might have a high level requests with multiple requested items and then each of those items having individual fulfillment tasks um so you would have like a parent of the request and then your child records would be items and then tasks and this is certainly something that we can accomplish will conf figure a couple of these examples later on in the instance itself but just to give you an idea of where this happens in the table configuration section of the project record you have the ability to add multiple tables and once you add a second table you have the opportunity to choose its parent and then the relationship of either child or reference um I do just want to point out uh that sometimes when you doing the analysis of the request process you don't always need the request table um more often than what I'm finding uh when working with customers is there's two Primary inefficiencies in the request process the approval step and bottlenecks and the Fulfillment tasks so if your approvals only live at the item level you may only need to use the requested item table as your parent and then the task table or the SE task table as your child um and just so you know there's a process mining use case series post on the community that details doing this type of analysis and how to configure this very specific project itself uh incident problem and change is another common example this happens to be a a reference example where the incident is the parent and the problem and change records would be linked to the instance via reference field I've also seen scenarios where the problems and changes are stored in um related tables uh on the instance so uh depending on where they're stored whether it be a a reference field on the record on the incident record or in related lists um we we might change whether Rec configuring it or linking it as a reference for a child so just something to be aware of this happens to be just a a reference slide of what the configuration for that use case might look like um so if you're going back and using these slides later I wanted to make sure I put this in as well the HR onboarding process is the third common example so with onboarding you're typically going to have a highlevel onboarding case and then a bunch of subtasks aligned to different activity sets and there's a a uniqueness to the onboarding example um there's some situations where you may not you may want to see your activity set value as the different activities in the parent map rather than let's say the different states that are on the onboarding case itself so in that situation you're going to have the option when you're configuring this type of project to say use the child table configurations as activities or activity values on your map there's a check box there for you to do this um and just to give you help get an understanding of how long it takes right if we check that box what that's going to allow us to do is get an understanding of how long it takes for employees to complete a full set of tasks in an activity set before moving to the next activity set of tasks um again because this is a slightly different well we'll configure um one of these examples in the demo portion of of the session as well all right so those are some of the more common examples other use cases include let's say incident and incident tasks change and change tasks and I've seen some customers look at linking Universal requests as the parent and then connecting those to both HR cases um and and incident and incidents out there so just kind of getting that view of how work is moving from those Universal requests and flowing into uh the appropriate workflow if you will okay now that we're starting to get a sense of what these Maps could look like and the types of connections I wanted to spend a little time on on how to interpret these types of maps uh within the individual processes themselves the analysis is the same if you were looking at them on your own in a single table map but you're also now going to have some arcs that are going from process to process or table to table and what questions can they help us answer and there's three primary questions they help us with the first question is how long does it take for a child activity or subprocess to begin uh so in the example on the slide we're we're we're looking at requested items as the the main process right or table and then the SE tasks as our subprocess and we have an arc going from the open State at the item level to the open State at the task level and what we see is that there are 187 that were uh tasks that were created and set to the open State while the parent item was in the open State and on average it took one second for those tasks to be created and opened um so that 1 second average there that's kind of an indication that is this is most likely already an automated process but if we saw a significant chunk of time there that may be an opportunity to automate or if we dug into that and we saw like a high average value and then we used the histogram to see that there was a couple of outliers like we want to understand like how those outliers even happened so in this next example you'd see we've got an arc that's going from the approval node in the requested item to the open state in the task meaning from an audit log perspective the task that created and went into the open state after the item was approved um in in this case there are 151 occurrences of that happening but what is interesting is the average duration is one week from the approval to the Fulfillment task um being created and and the work starting like why is that happening is there an opportunity to automate something there that we we haven't automated already uh so it's in these situations where we're looking for arcs that are going from parent to child processes where we're looking for situations to reclaim some of the loss time before work is being started right so sometimes there's inefficiencies in work getting created or or us even starting the work so it gives us opportunities there then of course there's the flip side of that right how long does it take for a parent process or parent process work to resume after the child activities have been completed um so if we look at the arc going from closed to complete and a task to to kind of close in the item we see we have 825 occurrences and on average it's taking about five hours for that to happen is there something we can do to reclaim some of those five five hours is there some sort of validation going on that we we need to be aware of what's Happening after the Fulfillment task completes is that something that we can automate or approve so this is another uh potential opportunity then the last question that these interprocess arcs help us answer is what's going on while one piece of work is in the in progress state or or maybe the hold on hold state right so for many workflows on the service now platform there's some sort of in progress or on hold state but knowing what is actually happening while something is in uh that in progress state is not always easy to come by so visualizing the subtest that are created while something goes into in progress or on hold can help us get a better picture of what's going on and where there might be bottlenecks or opportunities to improve now as we ran through these examples you might have been thinking some of those ins and outs numbers didn't add up like why were there some items that had approvals and some that didn't I and this is because in in these examples we mined across a bunch of requested items and they each could have had a different process and the reason I call this out um is just really more of a reminder that if you're looking to analyze requested items sometimes it's best to narrow your project down to one specific item to make the analysis a bit easier and and what your seeing in the data a little bit more consistent right um some customers I've seen tend to just they point every you know they point a project at all of the requested items and it's like okay you're going to see a little bit of a messy situation there it's useful in some cases to help you identify inefficient items um at a high level but uh you potentially could use reporting or performance analytics to do that right what we're trying to do is here once we identify an item that might be taking a longer time or have a longing fulfillment times let's just focus focusing on mining for that specific item and'll be easier to identify the bottlenecks in that process all right so now that we know how to interpret the arcs coming in and out uh the different process in and out of the different processes or tables on the map and the types of questions they help us answer what about interacting with the workbench there there's some things you need to be aware of when you're interacting with multi-dimensional maps versus when you're looking at a just a single table map so first uh when you open a multi-dimensional map the child table's automatically going to be collapsed um and when that table's collapsed you'll only see one Arc and that will give you the total connections between the tables you'll be able to then you'll be able to expand the child table and when you start expanding it you'll see all of the individual connection points between the the parent and child tables now in order to get a really good understanding of all the connections it's helpful to use the model options panel on the right hand side of the screen to expand the number of connections for each table like if a node is not showing um on one of the tables either of the tables you're you aren't going to see the connection so kind of for these multi-dimensional ones sometimes it's helpful to just expand uh the connections out to 100% on both that way you get a full picture before you start ning it back down now this is super important in terms of breakdowns and variation analysis um you're going to have breakdowns and variation analysis available for both parents and child tables but only the parent time uh parent table values can be used to filter the map so um child the child table breakdowns they'll be there to provide contextual information um and the routes will be there so the variation analysis but you won't be able to select them and filter the ma that's only going to be available at the parent level for breakdowns and variation analysis it's also important to point out that I point out that if you filter the parent table by the breakdown um the child tables will filter down to reflect that change as well so you're looking at some of the breakdowns at the parent you click on a box you hit apply and now it's going to narrow down just to show you the the child Rel child records that are related to that that breakdown value that you've selected uh so just know that that's that's how that interaction works you'll also be able to create transition filters across the different tables so this is super useful when digging into those use cases we ran through earlier so for example we can create a transition filter to n down and only show us situations where took longer than one day from the time the item was approved until the Fulfillment task was open so if you're interested in learning more about transition filters like this is the first time you're seeing a screen like this we have a whole Academy session out there on this I think it's Academy session 11 um isolating opportunities and mind data we really dig in and spend time on how to use these transition fil filters but for those that are familiar just know that you can apply these across the different tables uh that you've got inside of a multi-dimensional map um and of course if we can create cross table transition filters that means we can create rule-based finding definitions or Improvement opportunities that can be surfaced on the summary and insights page uh if you aren't familiar with finding definitions you should check out the last session we did um on automated findings or uh there's an earlier Academy session out there specifically focused on rule-based finding definitions but by creating a finding definition you allow the Improvement opportunity to automatically be generated and then present it on the summary and insights page when you remind the project in the future um so it's just like so we can do it in an ad hoc fashion using what we just saw on the prior screen here right cross table transition filters that's available in the workbench itself but if it's something that you're going to want to keep coming back to over and over and over again you might as well create a finding definition on the project uh so each time you mind um the the incidents or in this case incidents but the work that meets that criteria will automatically be surfaced um in your improvement opportunities list on the submarine insights page two last items before we jump into the instance um these are both Vancouver level enhancements so the first is that you have the ability to add an activity node on the parent table to show when child records are being attached uh this might be useful for for some to make the like just to make the interpretation of when a child record is being added in the process uh easier to read I find it useful also to to have this here if and use the repetition histogram on the Node to isolate situations where multiple child records were added uh to a single parent record um and it's also useful if you're trying to create certain transition filter scenarios out there so it's just an option that you have um to to add kind of instead of just relying on lines coming from different nodes on the the parent map to have a specific node on the parent map of when a child has gotten added a child record gets added or attached and then the last capability and for my friend out there who asked about the SLA breach use case yesterday this is the the thing that I was thinking of uh the last capability that's also something we added in Vancouver as most of you know we use the the timestamp in the audit logs to generate the process Maps now there's going to be certain situations where we may not have audit log in a table available to us uh but there might be some date timestamp information available on the record itself and we can use that to help in our analysis so for example uh we may not have breach time available to us in let's say the incident audit log that information lives on the task SLA table but we can add the task SLA table as a child and set the breach time as an activity uh to include on the process process map uh this will help us see where the breach took place in relation to the rest of the uh activities uh in the project or on the map um so when we use this capability the the non audit activity is visualized as an additional table on the map or an additional process on the map and then you can also add a child attached node to the parent to help with the readability of this situation so you can start to see uh where and where the breach actually occurred um very popular way to use this scenario is to create a project with assignment groups on as your activity on the parent and then add this breach so you can see which group had the ticket when the the breach actually occurred uh this is a specific as a number of customers asked for this capability um and it manifests itself as a multi-dimensional view so I figured i' I'd put it out here all right now let's jump into an instance and just look at some of those things that we just covered live so we'll come over here um and I've got my incidents with tasks for Academy project now really really simple as you would imagine creating demo data for multi-dimensional uh projects isn't the easiest so I've just created a couple of records moved them through the process so we can get a view of what this might look like and kind of make it easier for us to just getting comfortable uh consuming uh the map and what's going on inside of the map so I'm just going to jump over here to the analyst workbench and the first thing that you'll notice on this one is that that we've got incident and we've got some incident tests that's our child table here um and by default when we turn this on uh these child tasks it's going to be collapsed and if I click on this Arc here it'll give us information about the total number of parent records and the total number of children records that went through were connected here so that's the collapsed view now if we expand this right we're going to start to see some of the ins and outs so in this case here I've got my four that uh as they went into the in progress State um the next step in the process for these four was that the incident task assigned aligned to it went into the open State and on an average it took about 27 seconds for that to happen because I was moving these things through really really quickly and then the flip side here right I've got closed uh complete on this time and then I've got these four that moved into the next step of the process was things going into the resolve State on the parent instance incident and I've got this one here that's floating out here and it's kind of like what's going on with that and this is where I was saying earlier it's really helpful to expand out both of the tables initially uh just to get a sense of kind of where all the ins and outsa are so if I came here to incident and I expanded the connections what you're going to see right is that I start getting a better picture of what's actually going on in one of these scenarios here I have this one incident that went into the onh hold State and as it was in the on hold State that's when the incident task was created 22 seconds after it moved into on hold right and then this one closed and then the next thing that happened in the parent record is it moved out of the on hold State back into the in progress state so that's just um sometimes you're going to look at the maps and you're going to be like okay not all the data is there and one of the things that's really important for you to do is just expand the connections so you can start uh seeing those things a little bit better then on the breakdown side of the house right so we've got our breakdowns and right now we're set at the incident table watch what happens when I swap this over to the task table which is the child in this scenario I no longer have the ability to select them um and filter on them and I still have my different variants or routes here but again I don't have the ability to select and narrow them down if I came back to the parent table here I would have that ability and I can say well maybe I only care about these three that were categorized as software and I can focus in on just those three and see how they were moving through the complete process let's clear it all and bring it back let's take a look at that transition filter scenario so let's just come over here we click on our transitions we can bring it up and now we can say okay only show me situations in which the state in the parent was let's say on hold and then it was followed by in the child table incident task the state being open and then of course I could add my constraints of I only care about the situations where it took longer than five days to do that or less than one day right that's a a good one to do with approvals right so if I had approval and one table I only care about the situations where it took longer than two days from the approval something being approved until the work started happening on the in the task level like what's going on there can we reclaim some of that time can we automate that in some way and now what I can do is I can hit apply and this is going to narrow it down to just that one situation that we created in which something went from the on hold state or while while the parent was in the on hold State the incident task was created and set to the open state so that's kind of how those things link together and then of course right what we would do here is if we we like this and this was something that we wanted to see on a regular basis let's just clear this we could turn it into um a finding definition and then that finding definition would be available to us in our Improvement opportunities list right so I've got my on hold incident to open in task scenario here the one we just created that's going to be available to us every single time that we mine and what we could do is quickly come in here and say okay well for all those incidents that met that criteria I want to now view that in the workbench so you can create those finding definitions on these um cross table scenarios if you wanted to do that all right last thing here I think I got everything I wanted to show in this one uh let's jump over to this just so you can see what that child activity would look like so in this case here what we've done is we've created a project um and in this project we decided whenever we are linking to an incident task right we're going to add this incident task activity to the parent table so I've got this here now I can see all the five times every time an incident task was created all five here right it was attached and we linked it back and then what you're seeing here is you know we've got the two outgoing arcs so remember we had four that went from here to the resolve State and then that one that was going from in on hold and then the next thing that happened after the task that created it went into the in progress state but you'll see now that I've got this child activity here right I can start using the histogram to look for repetitions right so if I wanted to let's say isolate I don't have something that happened multiple times here but let's say I had a a parent incident that had multiple tasks that were assigned to it or attached to it you would start seeing in the histogram that I had two three four repetitions of incident tasks being attached here um so it's it's useful to do that type of analysis where you might have multiple um subprocesses or multiple tasks aligned to a parent table cool let's go create and I I see that I've got some questions coming in I'm going to I'll look at them in just a minute or two I want to create one of these projects so walk through the creation of the project and then while that's mining we'll have some time to to check the questions so if you've got questions on anything that we've covered so far so far now is a good time to put them in the Q&A uh because we we we'll likely have 30 seconds to a couple of minutes to answer them in a minute here so let's just build one of these multi-dimensional projects so we'll come over here we'll say create project and we'll just build that one that we were just looking at the incident in incident task so we'll say um Academy incident to incident task project and maybe we'll add a dashboard and we'll put this uh instant process mining dashboard on here and we hit save and then we'll come down here to our table configurations and the reason this is a related list is because we can now add multiple tables to this so we'll click on new and the first one we'll add is our our incident table and we'll say incident we'll come down here to our table list we'll find our incident table there it is and we'll say active is false of course because we want to look at the closed work um and then we'll just try to get those incidents that I've created recently um so I think let's just say they closed in the last week closed on this week and if I preview that should get the ones that that we generated and we'll hit save all right our activity definition here we'll just use State and one thing I'll point out you notice here activity definitions I've got two I got my field values from the parent table here's that date time value um so this is how we would do that SLA breach example but you'll notice here right now I don't have child activity because we only have one table when we we come back in a little bit we'll have another opportunity to add an activity in a second so we'll hit State as our activity so we're going to be looking at how the state is changing um through the incident process on the parent table here and then we'll go over to our breakdowns as you know you can have up to 10 of these we're just going to add one uh for this scenario here let's just say uh I think category will work yeah category will work I think I added that to these records great got our our parent table so now we'll go back to the project definition and we'll add our child table in this case it's going to be our incident tasks so we'll come back to the table configuration we'll hit new and we'll say incident task incident task great got it we also set these to active as false and I think we'll say closed on the last week this week close this week perfect preview shouldn't be too many of these yep there's the seven and now you notice here I've got this parent table configuration and if I look at this right I got my incident table that we configured um already we can link it here it's going to ask me the relationship reference or child so um the way that I always look at this if it's one to one it's usually like there's a magnifying glass on the in this case incident record and I've got one specific value you would choose reference if it's a situation where you could have one to many so an incident could have multiple tasks aligned to it would be in a related list connected to it you're probably going to choose child that's how I help reminder myself and then what field is going to contain that linkage and in this case the the linkage is going to be on the incident record or in number so we'll say okay we hit save and then we need to give this child table some activities and breakdowns as well in this case we're going to choose state but we could put assignment group or anything that we wanted here and we'll hit submit and then we'll give it one breakdown I think on these tasks I only added priority so we'll just quickly select priority and we'll say submit great now if I go back to the parent table configuration actually go I'll go to the project definition just so it's clear now what you'll see is we've got both incident and incident tasks as table configurations in this project now I want to just show you so we can if I wanted to add that child activity as a node an activity node on the parent map let's just go back into that incident table now and if I go to my AC ity definitions and say new I've got an option here to say child attached remember before we didn't have this we only had field values in date time we've got child attached now because we have a child table so we can come here and we can get our child table of incident task and we want to say that the incident task attached and that's going to get us that node on the parent map when we mine the data all right we should be good to go you know what I can do though let's go to the project definition we'll import this is something new for those that have not played around with Vancouver yet we want to import our process finding templates and our our finding definitions so we'll import those that'll get us those Improvement opportunities on the summary and insights page both rule-based and automated and now we can hit generate model um and for those that if this is your first time what happens at this point is it goes harvest the relevant audit log data from based on the criteria that we put inside of those conditions so like active as false and closed in the last week it then takes that data it packages up it passes it over to the machine learning infrastructure to do the heavy duty number crunching uh this usually takes anywhere from 30 seconds to a couple of minutes depending where we are in line for access to that machine learning infrastructure in this case we we move pretty quickly it then once it's done doing the number crunching it spits the Finish model back into the instance um and away we go um so here we go we've just generated that multi-dimensional map if we open up this workbench and this might take a second or two we've got our 11 records that we M with the different routes we've got our summary and insights page loading up here um and then let's just go over here to that process map and what you'll see is we've got our tasks collapsed here in this case there's seven unique child records but six unique parent records so we can see the difference there this is not the exact data that I had before you'll see that we've got our incident task activity here uh that child or so here where we started connecting the test let's just expand this um and you'll notice on this one I got a couple extras so there's my histogram so if I only wanted to focus in on the situation where we had an incident an incident task attached or more than one incident task attached to this parent incident I could focus on that and narrow it down in the histogram there and let's just expand this out a little bit so we can start seeing the data a little bit better and now we've got everything flowing through that that's what it takes to create a pretty basic um uh multi-dimensional process map right so we've got a par we just did a parent and child and we also saw how to add that child activity as a node on the parent now last one that I want to configure is let's I mentioned earlier that that HR life cycle event had a little bit of a uniqueness to it so we'll we'll go and we'll configure that one and then we'll we'll get to your questions uh so let's come over here and we'll say create new project we'll call this one Academy HR onboarding I never know if onboarding is hyphenated or not but we say table configuration and the first thing that we're going to do is we're going to choose our um HR life cycle case says our parent so we'll say HR life cycle case it's going to be our parent HR life cycle event case cool in this case here the values that we need to set in My Demo data is not all that clean here so and I don't have all that much but we'll just say the parent is empty in this case here and I just to make sure I get the records that are specifically for this process mining solution let's say they created by contains uh po make sure I got some data there all right got some records good we'll hit save now in this case here we're going to be using our activity sets right the the collection of subtests as our activity so I'm not going to add an activity right now because we're going to come back and do something special in a second or not a second probably two minutes or three minutes here we'll just add a breakdown though let's just say a assignment group and hit submit and now we'll go to the project definition and start adding some child tables and we'll add two in this case so we'll come here and we'll say new and we'll add two sets of tasks for different activity sets the interesting here this is something that we've not done earlier in this session is that they're all going to be coming from the HR task table but we're going to be using filter criteria to pick and choose kind of how we bucket these tasks so the first set that we're going to do is going to be our our pre-boarding tasks and we'll come here to our HR task table task table and then we're going to use our activity and do walk over to our Activity Set title value and just say hey I only care about the uh pre-boarding tasks let's see if I got that right yep there we go there's our pre-boarding tasks cool and now we got to choose our parent table configuration which is going to be those life cycle cases got it the relationship is going to be child because it could be multiple tasks Associated to um this and then we choose our parent as our reference task table and we hit okay and we say save now we'll give it a breakdown uh I'm going to say breakdown of let's just do assignment group again and hit submit and then let's do our activity definition of State because in the child tasks we want to have that activity definition of looking how things are moving through the different state values so let's come in here we say State and we hit submit cool we got our first one and now let's go back to the project definition and we'll add a second child table in this case for a different set of H our tasks we'll call them day one tasks I think they are so we'll say new these will be our day one tasks and the table is going to be exactly the same our HR test table but in this case we're just going to filter it and look at a different set of HR tasks and Link them as another uh Activity Set on the parent map so here we say Activity Set we come in here and we say title and I hope I remember this I think it's day one and we hit preview look at that I got some memor is not g yet and again we want to link this not to pre-boarding but to our our parent of HR life cycle case child and then link it on the parent reference here and say okay save cool activity definition one is going to be State again and then we'll use that assignment group got it and our breakdown definition now I think just so as I'm going through this you know one of the things that I'd like to call out and maybe I'm obiously a little bit biased here but you know this is a pretty complicated thing or complex scenario here we've got HR onboarding like parent cases child tasks all these different things and we want to try to process mine across all of that uh the the fact that we're going to be able to do this in in under five minutes um I think is pretty pretty impressive so here now let's go back now that we've got our two child um tables and our you know the the the different activity aligned to the different activity sets we can come back here and let's go to we don't need to add this activity it's actually on the project definition that magical checkbox that I talked about earlier so use child table configurations as activities so let's check this box right and now what's going to happen is instead of you know seeing State transfers in here we didn't actually add a an activity on our life cycle it's going to say we're going to see an activity for day one tests and an activity for pre-boarding tests so the two child tables are going to show as activities uh let's save this it's probably going to get mad at me that I'm not importing any process finding templates but that's okay yep got it I'm sure let's go and we'll this data um and this this again anywhere from 30 seconds to a couple of minutes so let's just take a look and see if there's any uh questions while this goes all right so got a question and I think we kind of addressed this in the slide so one of them is hey is there a use case uh around interpreting paths from incident table as the parent to task SLA as the child um and and we kind of covered that in one of the slides so that being able to link to the task SLA table to show the breach time um is definitely a use case that we see so I hope I hope that you know the the slide and the example that I showed there covered that um but I I definitely had somebody else ask a similar question so it this is something that maybe we do either a use case video for um or we come back and do a future session on that but I hope that's slide answered that question that that yes there is a use case for incident table and task SLA as the child um looking for that breach analysis now uh next one here so the question does this only work for workflows in service now or could this integrate with another system great question um so while we're doing a lot of the examples here on service now data today um and that's you know we're you know you saw you've seen how straightforward it is to do this for service now tables as of the Vancouver release of service now we give customers the ability to extract process data from let's say a a workday or that smarter smart recruiters example that I I showed at the very beginning um or if you have some process data in Oracle or Salesforce or wherever it might be if you can get us the audit log information or access to the audit log information so when I say audit log information what I mean is a unique identifier for the record whether it be an inight the CL ID a patient ID whatever it might be for the record an activity so something that's changing over time in the examples that we've used here today uh we focus mainly on state transfers or state changes as activities but you certainly could look at things like assignment group changings over time that's a very popular one uh if you can get us a unique identifier an activity and a timestamp value from any external system that's running a process for you um and we can feed that into service now uh imported in and apply our process mining solution to it so the answer is 100% yes uh so it's not just service now data it's data from other processes as well if you can help us get access to that so certainly would love to explore that use case with you for that person that asked that question um so somebody's asking about the the on hold example um so uh the the idea there is like how does that even work um so there's situations out there where like uh a uh somebody is working in incident and let's say they need to uh go out and get some additional information from a vendor uh sometimes what happens is they put the incident into an onh hold State and they give it an onh hold reason of let's say a waiting vendor and then maybe a task would get assigned to the person that's working with that vendor so there would be a subtask that would need to be worked while it's in that onh hold state of a waiting vendor um and we just want to get a sense of kind of hey what's going on in that you know in the subw workk that's going on there so that's a situation where you would have uh something being in on hold so in the parent table you've got your on hold step that's the last thing that happened in the parent table before the subtask was created or the task was created so you see the line at that point in the process when that subw workk was created the parent record would move back into an in progress State and that would be the next thing on the parent record that would happen so that's why you would see the arc going from on hold to the the child table and then the outgoing Arc from the child table to the in progress State I hope that helped clear that one up all right so this completed Mining and let's just take a look at what our HR onboarding example would look like we'll come over here to the workbench so remember we created parent uh table of the case and then two subtasks linked to it right or activity sets one pre-boarding and one day one tasks and then what you'll notice in here on the map is that those things are showing up the child tables are showing up at different activi activities and what we can see now is this uh volume and average duration going from the pre-boarding activity set to the day one tests what that is showing us is it takes the last time stamp let's say for a an HR case pre-boarding task um and then it kind starts to say okay there's the last time stamp there where's the the first Tim stamp and the day one tasks and what's the difference between those two um and then just like everything you saw us doing on the incident we can start expanding the out and obviously from an onboarding perspective we only did two sets of tasks here you could this could get pretty involved um as you start to look at your entire onboarding process in some cases you know some of these tasks we linked some HR case or HR tasks but you're going to have it tasks you're going to have maybe procurement tasks uh inside of uh an onboarding thing so being able to kind of Link together both it finance and all types of related tasks assigned to an HR life cycle case like this that's something that you can accomplish and get that end to-end visibility um using in platform process Mining and it's it's pretty straightforward as you you saw here today it might take once or twice to get used to it but once you you do it once or twice you get the gist of it you know in a matter of minutes um you you have some visibility into this process that you've never had before all right I know that we're going a little longer than we normally do here but I wanted you to see both of those things now any so we we addressed some questions are there any other questions out there uh any any questions about anything that you you saw here today anything just put them in the Q&A uh section we'll do our best to answer those and for those who asked questions um I hope that I I address them properly if you have a follow-up question please put those in as well all right while you're getting any last minute questions in I'll just come back here and um just know that you know you're not familiar with these things there's tons of resources out there above and beyond today for you to learn more about service now's process mining capabilities so if you go to the now learning site um there's a a process mining Essentials course it's about 90 minutes um 45 if you do it at two times speed that's sitting out there for you free just got to log in and sign up for the na learning site for those of you that are starting your process mining journey and you want to get IDE of like what goes into creating maybe a process mining program we've got a really good now on now white paper of how we've approached process mining here at service now and some of our initial findings uh there's a community Forum which I believe is probably how you found this session here today um but if you you're not familiar with it and you've got questions above and beyond what we covered go to that Community Forum site you can post your question we check it every single day we we try to answer things within a day when you post them there I check every single morning while I'm drinking my coffee and and get back to folks this is also where you'll find the other recordings of Prior process mining Academy sessions or those use case series videos that I've mentioned um so that's a good place for you to get familiar with um you're here so you know what the process mining Academy is but if you just happen to stumble upon this one class or session uh there's 20 other ones out there right now that are recorded um that go into different aspects of the of the solution um there's a use case Series so as of now I think there's eight use case videos we have out there so what these are 5 minute videos that talk about how to do things like SLA breach analysis or how to use process money to look for work that takes longer to route and resolve or how to use process mining to look for pingpong scenarios within your assignment groups and then lastly um there's a live on now session out there that just talks about getting started with service now in platform process mining Solutions so uh what plugins to turn on how to create your first project although you've seen some of that here today um it starts from the very beginning and walks you through the different building blocks what goes into our process mining solution all right last chance for romance last chance for questions let's see I see a few new ones that have come in let's take a look perfect and I see my friend who asked me about the SLA use case earlier on it's a hit awesome I can't wait to hear how it works for you um perfect got it cool awesome so most of the these these new things that came in were not questions they're just feedback that we we answered the questions which is great um and you know I will say for those that this if this happened to be your first exposure to service Nows process mining this is a very very Advanced topic I would say um so go back and check out some of the other ones getting started with this is super easy um many customers can get up and running in in the first hour that they've turned it on they find some interesting op observations and things that they potentially should be acting on from a process inefficiency perspective so um go and check some of the the earlier sessions as well I hope this uh for those that have been around for a little bit I hope this hit the mark with that um last thing I'll mention is that this is it for for this year the next session will be in the New Year we're targeting uh January 19th and right now uh we're targeting looking at using a session to focus on how do we apply process mining to custom applications that we've built with app Eng on the platform and then uh you know well we have to think up what the rest of the sessions for next year will be and just from an invite perspective this is the last webinar on this invite I'm going to try to figure out in Zoom how to just extend it and add these monthly sessions but if I screw it up uh you may get a new invite for a new series of webinars next year I hope that I don't screw it up because there's several hundred now that are kind of registered for this already and just kind of showing up on a regular basis but if I do uh you might get a new invite and I apologize for screwing that up with that thank you everyone we we'll see you in the new year appreciate your time today and marry mining