okay we're gonna go ahead and get started here and we still have some people that are trickling in but we'll go ahead and go over our logistics information that we have here so uh good morning good afternoon good evening uh welcome to platform analytics um if you're new here welcome if you've been here before uh thanks for coming back just real quickly like i said we'll just go over some logistical information and we'll get right into the uh presentation that we have today as usual this is for you so please take you know full advantage of it and you know ask any questions that you may have and you know the idea behind these are to you know bring you fresh ideas and better understanding of some things that are already out there and of course some practical guidance where we can to anything that is inside the platform uh as well so this will be recorded so um you can sit back and just watch you don't necessarily have to take notes because it will be available to you later um if you want to take notes that's great as well but it will be out in on this event and in the community and also we'll have it out on youtube as well and there will be there is a q a section so please if you have questions feel free to ask if you can put those in the q a um box itself right there rather than putting it in the chat we'd really appreciate that we do have uh other people outside of the pre presenters here that will answer questions as they can if questions come in that are relevant to this uh the topic and also the uh certain section that we're going over we will stop and ask those questions uh rather than waiting all the way to the end uh and at the end when we get to the um actual q a section if we can get to any questions that are unrelated to the content going over today we'll try as hard as we can to get to those but we will answer the questions related to the content uh first if we're not able to get to those then you definitely can put a question into the community and tag any one of us and we'll help you with that so today we're going to be talking about conversational analytics uh in the dashboard and everything that comes with that i am the host thomas davis and today we have on uh two presenters we have victor and then we also have on shesh who will be going over uh like i said conversational analytics and some really uh in-depth information about that so if you're unfamiliar with what it is they'll show you what the dashboard is and then also talk to you about all the things that come in there and also go over some uh how to's of how you can create some things that go on that uh as well as well as us we also have some panelists on today so adam stout and tara romero and dan kane who are all on my team are here as well they will answer questions in the q a um if they can and uh if they need to chime in they definitely will as we get further along into it um so again uh with that i'm going to turn this over to victor and we're going to go ahead and get this free stage presentation started victor you there yes i am thank you thomas thank you thank you for the introduction uh like uh he mentioned i'm an outdoor product manager for conversational interfaces and what that essentially means for those who are unfamiliar is uh uh you know sasha kant and i are product managers for virtual agent uh agent chat and advanced work assignment and today we want to share the um the conversational analytics dashboard which is essentially a dashboard to measure the effectiveness and the usage of your virtual agents so for those of you who are using the virtual agent you can actually measure and communicate uh the the roi that you get with the virtual agent to your stakeholders or to your team okay generally much like your every other academy we'll do a little bit of a slide slide deck overview and then we'll jump right into the instance with uh with sesh so an overview again of the endless dashboard uh in the it is new as of the quebec release and then in the rome release uh we do have uh new features which uh coincided with the store app released in the fall so the analytics dashboard is a store application and some new features we released is a deflection tracker which we're going to go over in a moment funnel builder and funnel widgets so if you know about funnel graphs and you know kind of get a sense what that's about again we're going to go over that in a bit uh saved views for conversations and users you can save a view or save a dashboard that you like seeing and then we have new metrics for one of our features issue auto resolution so you're going to hear the word deflection a lot uh today and um you know what does that mean or what does that word mean in this context well deflection means that a user a virtual agent end user has had their issue resolved without creating a ticket or interacting with a live agent so that is the goal of virtu of the virtual agent is to automate your most common uh end user tasks and questions and issues without getting a live agent involved that way the live agent they can concentrate more strategic and more important uh goals and and issues so uh that's our definition right but you uh can also create your own definition so you can create your own custom deflections again which we'll go over to uh in a moment but we do try to provide you with as much out of box information as possible and with the goal of right being able to handle more than 50 of the most common call drivers across it hr and customer service management again freeing up your support staff to focus on actual value added tasks so that i'm going to hand it over to uh cheshire khan he's going to give you a brief uh overview of the conversational dashboard for those of you who aren't aware and then come back to me we'll do some more slides and we'll do some back and forth so uh with that uh shesh are you there and do you want to take control of the screen share yes yes victor thank you and uh good morning good afternoon good evening to all the people who joined the call uh myself i'm part of the product management team at the conversational interface system or a platform so before me getting into the deflection so i would like to start with the introduction of conversational analytics uh dashboards let me quickly share my screen so i am going to uh show you the virtual agent dashboard so this is the dashboard you know home page of the conversation analytics dashboard where you have the complete visibility of of the or the performance of the virtual agent deployed inside the organization so to get to this page you just need to search for uh dashboard so so under the conversational interfaces we have the virtual agent dashboard you just click on that and you come you come to the uh dish overview page of the virtual agent analytics where we pretty much sum the uh the overview page basically tells the about the first three important metrics on the deflections and uh the amount of deflections the contributed by the virtual agent inside the organization and we also divide this number across the different patterns uh we come to the deflection patterns in the in the next slide or the the next topic actually and we also show the kb articles that were shown or that were helped to deflect these number of tickets and i quickly run through some of the metrics to give you the complete visibility of the dashboard virtualization metrics that we represent here we talk about the active users the number of conversations and the success rate of virtual agent and topic flows completed so these two metrics are also equally important va success indicates uh the uh the conversations completed without escalation to the live agent and this is an important metric as well so that you came to know you come to know what's the percentage of success for my virtualization and here you can clearly see that uh 64 nearly 60 percent of the time virtual agent is successful and 40 of the time it goes to the live agent there could be multiple reasons for the transfer to the live agent that you can come and dig into the conversations data from here as well and uh 82 of the topic flows completed means rest of the 20 percent people are not completing the topics maybe is the issue with it the way topic got developed developed or some issues with the user is not able to uh provide enough data there will be various reasons or some of the integration skulls might be failing so this is also an important measure for the success of the venture agent along with the other deflection metrics that are represented there and the overview page also talks about the when it comes to top categories right we we all divide the topics across uh multiple categories like could be idsm topics or you know csm or employee related workflow employee related topics you can divide the topics across multiple categories and then you can see category wise or how much how many of them are completed how many of the income plate and similarly for the topics as well uh uh how many times this topic was completed you are looking at the software access as a topic that got accessed almost 64 number of times for this time region so i just selected the top let's say for last three months um this topic got initiated or executed 64 number of times and you know you can see the order and similarly you can also see the incoming completed conveyor topics as well so as you could see here food menu uh is the most incomplete topic there is something wrong with the way it was developed some sort of you know this matrix gives you the you know some visibility into how your topics got initiated and completed as well and similarly we also divide the transaction count across multiple channels that were configured here we have three channels configured web client and slack and teams for this speculation and the breakdown of the conversations across these channels and similarly we also represent the conversation in state you know how the uh the number of conversations ending uh for this conversation instead so conversations could be ended by in various uh end states some of them it could be it could have completed uh successfully our agent could have closed it or user could have closed it some conversations might be timed out like that we represent the various end states and then the number of times that that state got occurred so this is the overview page and there's a very consolidated view of how the overall performance of the dashboard which will be much legend coming to the other uh page where we give the usage level metrics so this depicts uh you know how many times the va was only initiated how many times the request to live agent was requested how many times the agent only initiated transfer conversations are there and a the average conversation time frame here [Music] is also depicted for the time period here and the language uh distribution how many times the spanish of [Music] conversations initiated versus english and the notifications actionable notification sent and how many of them expired what says the topics that got initiated by the notifications we can add few more metrics to this edge but pretty much this is uh fairly what we have for the rom up to row version or san diego version new recent release the conversations tab is uh is is the an important workspace where we give the data of the conversations data related to the exact conversations between the user and the virtual agent and the user and the live agent as well so let's say this conversation got started uh for this time frame that we present here and we also give uh for this timeline or timestamp it got started the start time and end time and the user id we represent and you can come to the transcript section here where you can see the the topic the exact conversation that happened between the user and the faculty agent here as you could see here uh i was asking about some sort of topic context as my um the search term uh so the topic name and then is asking part is asking for some entering some keyword and then user is giving some sort of uh input to that so as a result it was showing some sort of data from the bot or from the knowledge or search so you can tend to see the exact conversation here as part of the conversations so you can select any conversation and then you know happen to you can see the conversation the exact conversation between the these two parties and also you have the editors also here where you can uh pretty well uh dig down deep diving into the specific channel wise or time-wise or you can filter the data as per your requirement and we also have user stamp which is equally powerful capability where you can see the all the active users and their conversations with the system and we have the topic conversation topics tab where we give the detailed analytics on the topics uh inside their virtual agent so what are the worst performing topics what are the best best performing topics as you could see here the software access stock access topic got initiated or executed most of the time in terms of the number of times they are executed for the given period of time and the the spokes that are in you know invoked as part of the uh the topic execution as you could see here we actual agent apa got initiated most of the time uh compared with the other spokes as part of the topic execution and we also talked about the topic level uh matrix and how percentage of the time they were initiated incompleted and the live agent request for for those topics and we also have the topic category level information detailed information you can pretty well choose the topic that you wanted to look at for idsm uh issues let's say you can till then get down into the talk conversation end statewide how many of them have been transferred live agent request usage trend and the channel distribution [Music] and you can also drill down to a topic level let's say you wanted to see the detailed matrix for the software access topic you can tend to uh download those data usages you know order given period of time how the trend has been so as you can see here in the 10 month of jan it has actually is a good spike in the month of jan it got more than 40 number of times it got executed how many times the topic got completed how many times it was incomplete and uh how many times the the live agent request uh was initiated so as you could see here there are 65 number of times the uh it was not initiated sorry here as you could see here number of uh sorry number of times it got initiated around 465 percent of the time it got people requested for the live agent when they are here though the completion was is good but equally people are requesting for live agent transfer when they come to this topic and you can also see this is also an important metric uh way as part of the topics here you people can see what was the last access the node inside this topic so typically people completely and come to the end node inside the topics for number of times and some of the people just stopped at here search one time and most of the time they completed 60 out of 65 number of times 64 times people visited the last node and only one time people visited stop radio ti search node itself i think pretty much is a detailed uh information at the topic level and we also represent the details at the spoke level as well this section particularly talks about all the spokes that are used inside the topics here uh coming to the next rap nlu predictions this is a i can say this is a [Music] basic stuff we are we are trying to improvise the per model analytics where we want to represent how many times the nlu was correct how many times nlu was incorrect and how many times nlu was not able to match the user request to any of the intents inside the model so you get the distribution of correct incorrect versus unmatched percentage for all the topic for all the nelu models deployed in the system and the graph here pretty much depicts you know how many times intent got auto selected how many times the uh the intent was correctly detected intent and content was wrongly detected so we just represent that trend here and uh some sort of you know we highlight what was the correct predictions and the incorrect predictions and auto selected predictions auto selected predictions are nothing but you the confidence score was pretty much high and nlu went ahead directly executing these intents without asking for the confirmation from the user so we just highlight the top five things here uh for each category and uh and in the time the near releases this page will be completely revamped and probably more insights on the nlp performance coming to the next tab user search metrics this is equally powerful capability that got introduced as part of san diego release where we are depicting the uh um the user search matrix is something where we wanted to throw the analytics on the what sort of content is being shared to the user and how good is that content at the end of the day so people are asking for this query and we showed something for that user did not click any of the results that were given so it pretty much all the links are the information that was shared to the user we give the analytics on this information you know how many times there the queries resulted with no clicks you user asked for something but we showed something we delivered something but user did not click on any of these results and similarly uh we also represent the queries where there are no results at all the user asked for something but virtual agent was not good enough in terms of identifying the relevant data to the user so for that it might demand the the creation of new knowledge articles these insights will really help the admin to know the quality of the data that is show to the user or share to the user and the effectiveness of the data that is shared to the users coming to the next two sections we give the trending queries and the trending content which is the which is mostly used uh mostly asked query and mostly shared content with high or click percentage of the content so this way we we show we these are the four sections that we delivered the content use content usefulness to the uh administrator and the content could be coming from ai search or contextual search or even the content might be coming from the topics itself sometimes the topic developers gives the links inside the topic itself that also we can you know we attract the effectiveness of the links that were shared from the topics so this is sorry i the want to give you a heads up for about 20 minutes and just doing a quick time check yep thanks victor yeah i will be summing up the other sections and uh the custom events page where you can build custom uh events and then get the detailed metrics from the uh the custom events that you build here the i search results displayed is the custom event that i built and then i see the data trend how this is occurring over the period of time and these are the custom metrics that you build as part of the custom events where you build in the title document table description data source and all and the another important metric uh important feature that got delivered as part of the rom was a funnels concept where it helps you to identify where users are dropping inside the conversation so one of the challenge most of the admin space is that people start the conversations and then leave in between and we need to identify which type in the conversation people are dropping and funnels are going to help you to identify that step in the conversation and we will get into the funnels as well as part of the discussion and the next section is on the issue order resolution where we sum up the metrics uh once the issue order resolution is deployed and we give the detailed matrix on the intents that match with the topics and intents without matching topics and all maybe i just leave it here [Music] so let's quickly get into our topic uh for today deflections and the funnels so as you could see here right uh in the first section we have uh the deflection metrics here for this time period va contributed around 46 deflections and those deflections were categorized for these deflection patterns so deflection patterns or vaself resolving and pa password reset pattern va search reserve pattern and you can these these deflection metrics are per configuration will come to know what is the configuration as well uh you can switch between these configurations and this matrix will change this for this software deflection configuration there were around 23 deflections and those patterns of these 23 were distributed for one deflection pattern and the kb breakdown here since there are no kb shown for this though this style is empty and as uh as briefly given by the definition right the definition deflection vector has given the definition of deflection and uh deflection matrix uh basically helps you to understand the the um how much of you know deflections contributed by the virtual agent to the organization so that's the critical matrix that you know helps you to justify your investment inside the virtual agent so anything that needs to be tracked first needs to be measured the way we measured the deflections is we have introduced the new topic block called reflection topic block that that needs to be inserted in inside the topic topic flow when you feel that okay at this point as you can see in that in the image right in the downside uh [Music] the image in the flow workflow here uh the deflection example is the top note topic block here so as per of this execution you know the deflection counts gets incremented so these out of the box deflection uh will be executed when the user control goes through that way so let's say once you have reached certain uh journey inside the topic as a topic developer you need to decide okay when the user reaches this position most of the job is done you may want to want to count the deflection so that's where you need to insert this new topic block that marks the deflection as per the topic exhibition okay so the total count of execution or you know execution count of deflection topic blocks gives the total number of deflections inside the maturation for the given time frame okay now coming to the deflection patterns right we have four out of the box uh deflections patterns uh first one is va self resolving va triads and created va intercept and resolved and vs search served so coming to the definitions of these patterns right uh vself resolving something where va was able to resolve uh user requests before uh you know you know user needs to go to the live agent or some sort of manual introduction let's say password reset examples right and you may want to uh simply automate that stuff let's say people come and ask for uh i need to reset my password and then you know evacuation to just ask for the information and does the uh executes the workflow which va results on its own so in this case when when you are doing some self-resolving stuff right where bot is actually executing certain tasks avoiding the manual intervention then you can consider putting your deflection into this category coming to these via trashed and created so where this section where you can you know put your deflections to this category when va is trashing or collecting certain uh actionable information and then creating the instant articles or creating the service request to the right uh assignment group where we avoided back and forth between the uh support person and an end user va was able to trash the required or actionable information much ahead avoiding manual back and forth between these information um exchanges and it was successfully able to create certain requests and assigning to the right assignment group in this case also you can consider this as a reflection and the third category is the va intercept and dissolved in this case uh let's say there was some request and it got created inside the system va goes and intercepts all the requests inside the in in your system and when there is a matching intent or matching topic that we can do it goes ahead and then i know engages with the user that hey i know you have asked for this request i can help you out and uh or you know is it okay to let for me to solve this request and uh va intercepts and resolves this case for these number of cases right you can also include uh if a topic is actually catering to those sort of requests also you can count you can map this deflection to this category as well and the fourth one is va search sometimes uh v8 does the knowledge search and serves the required information to the user so for this case also you can put this category and typically the knowledge breakdown the section in the in the dashboard right in the page visual that is event guide the kb breakdown uh category basically it comes to the va server category then you may ask okay out of the four deflection patterns is there any way that i can create my own pattern let's say all your password reset cases right you may want to create certain custom pattern okay all these software access patterns and you you may want to create new pattern and then you may want to distribute that count the deflection to that pattern that's pretty much possible let me quickly get into the demo of this how this would look like for this i have to shade my screen let me take the one example uh one topic let me show how the topic block gets executed was built so as you can see here the topic right we are asking for the application name for the user so user selects one application out of these two choices sap and onedrive and if the decision is uh the user input is one drive we come to this path and then we uh just execute this flow for this example i just put down this message like one drive access for flow initiated and uh once this is done and this is where we wanted to mark the deflection okay my job is here done most of the uh the part is done then before coming to the end i put this deflection okay when the user comes to this path it's more or less the job is done i would want to you know in you know increment the deflection count so i just put this deflection uh a block here and drag and drop here and if you just select hit here and you can provide the configuration so for this i may want to create the new deflection configuration software i am not choosing the va default configuration whatever is given rather i created a new configuration under software i may want to put the deflection pattern as new deflection pattern software access as the new deflection pattern here i created this is a custom deflection pattern i created and [Music] that's it so when you execute this topic the count gets incremented let's see in the deflection software configuration the deflection count is 23 and let's try to execute this topic and see if this count gets incremented currently what we have is 23 number of deflections so i would want to execute the software access topic it is asking for which application you need access to i'm going to select onedrive and yes by this time the topic flow could have been executed let me come and then see this dashboard is reflecting there yeah earlier it was 23 and now it is 24. and this is the pattern it is distributing the logic distributing the count so that's how you need to uh you know use the topic block and when it comes to defining the configurations right where do you define custom configurations and all right you can do this in the settings of settings page of the virtual agent so we come to the conversational analytics settings page here and you just come to the virtual agent settings when you come here we have the deflection matrix section so you just come and then you know see what are all the deflection configurations that you have defined so i have used the deflection configuration software under this and uh i may want to create a new pattern let's say i want to create new pattern all the sap access pattern and you may want to define what is the outcome is it going to be the confirmed deflection or no deflection or potential deflection so for these cases let me say it's a confirmed deflection and you may want to also associate some of the primary table uh in to which you wanted to associate this deflection to uh if it is incidents you can have the uh you know associate the inclined incidence table and it's not mandatory though so i just leave this and yeah i just create a new sap access pattern and i may want to come to the designer and uh one second let me come to the designer and edit this topic i came to the designer so let me edit the topic uh deflection node pattern to the newly created pattern here so under the deflection software x software configuration let me choose to sap flow that is just created so sap access pattern so let me save this and publish the data to the model so that we get the new definition let me go ahead and execute the pattern node topic again here [Music] um [Music] the topic got executed so let me refresh this page it should be 25 at the same time the title also should be reflected here yes you can see here right we earlier it was 24 now it is 25 and the software pattern deflection pattern software access was 24 and the new pattern is after sap access pattern for that the deflection was contributed to one so that's it on the uh deflection patterns and i would like to quickly jump on to the uh funnels feature uh where the main goal uh of the funnel feature is to identify where users are dropping in the conversation steps so the conversation can have multiple topics and various uh steps inside the topics as well so let's try to see how the funnel creating a funnel would look like before let me show you uh what how your funnel would look like right so when you create a funnel right this how it would look like so it each bar here represents the number of users at that particular step inside the funnel so here there are 100 users and when they start these their top they are starting the password reset with no sso topic and then they are entering into the note step two where they are required to provide email address and when they come to this step only eighty percent of eighty users eighty percent of the users coming who are here and twenty percent of the people are dropping at this point in the in the topic or topic or conversation so you can see that okay people might be having some challenge with providing the email uh email address at this step inside the conversation and the third step is uh like the node answer any security question where the bot might be challenging with some of the security question and this is where it would instruct the user we'll have that number of users at this step so looking at these steps right you may want to see okay if here users are starting and users might be ending here and you may want to see the breakdown of at each step which step is actually causing more user drops off that's the whole intent why the funnel is going to be useful and you can define maximum of 10 steps inside the funnel and the three tiles here talks about the percentage of users who made through all steps defined here and let's say your goal might be okay when i'm starting this topic i want to track each and every flow inside the topic for let's say for this topic i developed for this topic i developed right of of all the users who started this topic software access topic and i may particularly wanted to see how many people came to this flow and finished it okay that if that is my goal i can build a funnel for one drive i can build a funnel for sap flow and see the overall breakdown of it or you may also want to see or i want to know all the users who came to uh access this software access flow and then they uh entered into this node application asking for the application name and they selected one drive application and then before ending right they may say that they before ending the conversation they could have given they could have asked for live agent request that could be your case you may want to see how many people access the one drive and then how many people went to live agent okay so that could also be one area where you may want to build the funnel right so this is where uh you know the use case where you you can build the funnels so uh yeah coming to the second step right uh here we depict the number of users who may drop all steps out of 100 users who started here 10 users actually made to all steps which means that 90 of the people dropped 90 users dropped from one step one to the last step okay and we also highlight okay what is the biggest step for user drops step one to step there is a maximum loss which gives an indication to that mean okay there is something wrong at step two right which might be causing more user drops off right so as an admin you may want to focus this strap this step and then try to improve that flow to wider bad user experience so let's quickly go ahead and then build one funnel so my goal is to know the list of all users who acts who traverse to this flow let's say uh one day flow i want to see how many people went through this flow that let's try to build some simple uh use case funnel one drive users you may want to see one drive users flow okay and you may want to see of all the active conversations or user can start from any conversation they could be coming from any conversation and they accessed uh the topic called software access topic here so which is a standard topic and let's see i am getting the software access this is my step one or step two okay and that step three is and i particularly interested to know the users who came to this path okay this path name is onedrive law okay we have the name here one drive flow let's try to put that flow inside the okay so that's next up is okay people who are accessing software access topic you may particularly want to drill down to the list of users who came to one drive flow one try one drive floor okay and let's try to quickly uh see this pattern so let's might take some time sometime it's going to take some time so let's uh see the other areas so in this example right we have the list of already created funnels here where you you just see here right one drive users we just created this funnel and you just need to click on this and then the visualization would start appearing so for the past three months right uh from 11th of uh january so november 2021 first to the uh the date march 1st last three months of all 22 active conversations or sorry 22 active users three users only initiated the topic software access topic out of all three users nobody came to onedrive flow that's the uh output we got here and this step one to step two is the biggest dropping point here that's how the computation was done on this funnel um i think we reached to the details on these two topics i would put a pause on the demo portion and then if we have any questions i'm very good to take great thanks sash so i hope this uh was a great overview of the conversational analytics dashboard and all of his abilities we do have a poll for you so if if our fellow host could show that poll we are interested in the level of your virtual agent implementation maturity uh while you fill out the poll i did want to highlight a question that was asked that that bear is repeating um we so this dashboard that you saw today was a replacement or an upgrade from the uh performance analytics or platform analytics for virtual agent that was delivered in pre-quebec we are no longer um building that out so uh we do recommend if you're in a quebec or later instance that you use this dashboard that you saw today not the performance analytics or platform analytics a dashboard for va that came perhaps in an orlando or paris release while we also while we do the poll if it's all right i'm gonna share uh i'll just share our our question or q a slide uh but also has some links on there let me do this so yeah so um you know we are again uh product managers for conversational interfaces uh virtual agent does have his own communities so feel free to uh uh stop on by and check it out the link is in um i think thomas is pasting in the chat otherwise can you can see the link on the slide we also have our own virtual agent and nlu academy as well which goes every other tuesday uh yesterday we just did one on enabling video and virtual agents so you can see we have a link for that as well in the slides there uh we're also open to any other questions you have so feel free to put them in the q a box while you fill out the poll okay great thank you victor and shesh so let's see we have some some questions coming out i don't know that this one is um directly related to the content so let's just see here if we have more um [Music] come in but really good information uh especially if you're someone that has never seen this or used it or you've definitely been asked to do some uh deflection information on you know anything that's coming in via virtual agent or other wherever else this is you know just really good information so i appreciate the two of you uh spending some time with us today and giving us that so uh let me go ahead and ask that or adam is going to go ahead and type that on adam unless you want to do it out loud we can do it out loud while we're waiting uh so the question that came in was about the number the limit on the number of elements for a breakdown in performance analytics uh and how it comes back how we compare to reporting specifically as well um i believe and if anybody wants to correct me i believe the out of the box limit for the number of breakdown elements is is it a thousand it's a thousand five thousand ten thousand it's one of those uh it's certainly not a hundred you can go above a hundred um there are restrictions for reasons why why we have a limit there and the base reason is is we materialize the scores and we store the the values of the elements um so at some point in time you just have reporting um and we're not getting the value out of the the kpis that we're driving so there is a limit and the more elements you have in your breakdown that can it can cause performance issues i like to keep it under a thousand um as long as it's under ten thousand you're generally okay but a thousand is better and a hundred's even better than that and if you are over a hundred you wanna make sure you're using elements filters um you're using the widgets that are using the personalized visualizations a human can't browse a thousand things right maybe a hundred i don't know what the number what the number is maybe you can scan over a hundred things but nobody's gonna scan through ten pages of 10 of 10 things and certainly not 100 pages of 10 things so we want to focus that up if we can hierarchies are really useful for this as well where we can roll up those uh those values into more meaningful elements there's a great lab on that i think dan i think it's your lab on now learning about hierarchy so i'd recommend doing that i know we've talked about it in some of our previous academy sessions as well but dan did a great a great lab on it and how to kind of build those things up the limits there for a reason and if you really are just reporting and i need to do something like all the breakdown by assign two you're better off in reporting generally those aren't great kpis you can do it and um but you know you want to understand what you're doing and breaking down by group might be better uh to do as well so yeah i want to say there is a property that there is a property for analytics hub itself that says like for analytics hub to visualize the there's the yeah yeah there are there's the number of elements that will be displayed yeah there's 100 but then there's also one that's analytics hub too you're right you're right because analytics sub is going to is going to show a chart so i'm sure there is a limit there as well for kpi details and analytics hub but that can be um that limit's not a collection limit that's a display and then when you search so if i search for atom it'll show me mine but if i don't put a filter on then it will show up it'll show it won't show anything because it's not going to show a thousand lines on the chart so that's where it comes to be really important to make sure you're using uh those those elements filters or personal visual uh personalized version visualizations in your widgets so that there might be a thousand but it's only showing me the five people for my team um so a lot there are except there's some other sessions on that the hierarchy one's really important look at the other tools to make it useful um but uh yeah if you've got to go from 100 to 200 you want to go to a thousand that's okay uh be conscious of what you're doing you may want to watch some of those other sessions about things you want to consider but if you're going to go to 100 000 don't you're just not going to want to it's going to cause lots of other headaches for you looks like we got another question about va as well yeah i think victor just uh just answered so the question was uh what rights are needed for virtual agent and um you shared pro and enterprise that was that was related to that question right victor yes okay so yeah so if we have va in use are we likely licensed for conversational analytics application i'll check ourselves but i thought i'd ask before i get too excited the answer that is yes you have the analytics dashboard if you have the virtual agent entitlement basically if you have itsm hr or csm pro or enterprise you have everything that we showed here today on along with virtual agent awesome thank you and just looking at the poll here real quick i think it you know 68 percent have not implemented so um this is a really good uh academy session for you know for that with with also some in-depth in-depth information that was shared so hopefully all of you that um said not implemented yet unimplemented yet will hopefully uh change that after this academy session so uh really good information and um a victor i don't know if you want to speak to the poll anymore or uh just that um is it possible to show the results or we can do that offline but yeah like i thought it was everybody was seeing it there you go let's see okay yeah like like was mentioned we definitely invite you to it certainly if you're entitled for it to try it out again visit our community or our product docs to get started with the virtual agent or or to try it out come in the san diego release you saw a preview of this in in cheshire's exercise we do have a a new conversational interface home page that basically gives you guide guided step by step um including recommended recommended steps on how to uh turn on your virtual agent very quickly very easily and get them you know get fast time to value so yeah again check it out and you know with the dashboard you can definitely show your stakeholders the value usage and effectiveness of that virtual religion that's that's the goal so um yeah i know that's all i've got uh again if you have any questions uh feel free to visit our community i think there's the one maybe it's related question looks like there but saying if we have va in use are we likely licensed for uh this uh for this conversational analytics application yeah we we answered that right before we started talking about the poll and he was saying that yes it is it is there for them absolutely yeah i think they kind of go hand in hand there right sorry yeah yeah that one and uh you know there's a new comment uh by uh mr rogers there um as you know sesh can correct me if i'm wrong but in general we have a parity if not more than parity between the existing platform analytics dashboards virtual agent and what you saw here today this is simply uh migrating some of those dashboards over to a technology that can kind of like you know track the topic node to node action more and then we build even more stuff on top of that so there should not there should not be a parity issue as it relates to to virtual agent if you're referring to the other questions then or the other uh pa dashboards then that maybe that may be right yeah there's other there's other pa functionality that the classic dashboard still has still happening but we can switch and use both right so if i'm i can still use my standard dashboards for all my itsm reporting on my hr reporting everything right and then when i want to dig into conversational analytics i can use that module and then use the this dashboard it's not an all or nothing right i don't i don't if you're focused on conversational interfaces or a va use this dashboard i mean at least use it in conjunction with that with this one for this purpose right um i see another question by uh by scott here do the analytics work with third-party producing iframe good question i believe the answer to that is yes as a channel it'll still show up as a web um conversation right because we only have web mobile or team slack etc so it won't show you whether it's like um internal external but uh you can still use again either platform analytics or some other or you know you create a custom event to say hey this came from a portal that was not service portal employee center etcetera because and then by process illumination you can kind of kind of figure out whether or not this was this was a web conversation that came from an external portal okay raised hand here too mike is he wanna i know we're just right up at the top of the hour but dudes we want to make uh unmute mike here real quick and see he had a raised hand go ahead mike or maybe he didn't intend to raise his hand i gave him permission maybe he was just hot air high fiving us okay air high five back yeah so um so real quick just a week because we're at the top of the hour uh we've got about a minute left so so mike if you do have a question you put that in the community on the event and then you know just tag myself and victor and we'll make sure that we get you an answer to which whatever you're looking for so quickly um just kind of go through some other things real quick just make sure that you're checking out the community uh great information there not only just the academy sessions that we have here now but also you can ask questions and and get information around getting started and how you hit the ground running and you know just really good information there in the community hopefully all of you are there um if not you should you should get there and get an account um all of the uh academy sessions are here we have a an academy page so all of the future stuff is there along with any previous that we have and we have that broke down by year and all you got to do is uh click on any one of those and you can get to that information and it is also out on youtube so this information will be on youtube later the event will be updated later with this information as well uh and don't forget about now learning not only for platform analytics but also like victor uh said in the chat with this sub subtile uh plug which was good that there is virtual agent uh training that is there uh as well so take advantage of now learning get you an account and get all the information that you can there uh and also some other academy sessions that you uh might be interested in create a workflow and platform academy uh mobile academy mobile app academy and also virtual agent academy which victor already talked to uh led to so the next time that we meet in two weeks we'll be talking about what's new in san diego analytics features we'll unpack all of that that is a two-part session because of the amount of stuff that's coming out so join us in two weeks for that and with that thank you and uh we are glad that you joined we look forward to you joining every week and if there's anything else that you need find us in the community and hit us up and we'll go from there so thank you again you