[Music] thank you [Music] thank you [Music] hi welcome everyone to life coding happy hour so if you are new to our show this is our this is our show that is not meant to be polished it's not meant to be scripted it's all about us just trying to get something done in servicenow so welcome to live coding happy hour is brought to you by the servicenow developer program we want unwind at the end of the long work week and I've I've definitely had a long work week this week so we drink we try to build something new and um we just try to show you the real development process we try to show you the errors along the way all the debugging that comes with being a real developer and all that stuff that comes with our actual jobs we are in the middle of our Utah release content so if print off can point over your shoulder to the logo yep there he goes uh that if you see that logo in the top Corner that means we are in the Utah release and so now you can kind of visually see all the different content that we are doing for Utah and on top of that if you go to devlink.sn Utah you will see all the stuff that we've been doing for the Utah release and it's a lot we've been doing this for weeks now and we're getting to the tail end we only have a couple more weeks of content for Utah specifically before we start unwinding or winding down from the Utah content that we have planned uh be sure to check that out though and then before we jump into our content let's do some introductions uh first off hi everyone My Name is Earl Duque I am a developer advocate here at servicenow but before servicenow I was in higher ed and uh jumping around a bunch of customers I did servicenow work for three different customers uh since 2017 before I joined servicenow in 2021 I think I'm I'm about to hit two years I think that's what it is go ahead and introduce yourself hey everyone my name is pranav I'm a senior developer advocate here at servicenow and I can't say I'm quite I'm new to servicenow now it's like already been four months before joining service now I was working with a customer and was helping them with implementation and before that I was working with different clients uh and helped them with providing them with the Consulting so that's about me I'll pass it to like hey everybody really excited to be there my name is loik Sanchez I'm an outbound product manager on the platform AI team ad service now I've been in that role for about a year and before that I was in the expert services so as a servicenow developer for customers also at servicenow and I also maybe you know me from the AI Academy show that I host on the servicenow community where we talk about all the the AI products that we have on the platform and there is a link in the description to that Academy that look is just talking about right now so make sure to check out that description if you just if you've been waiting for the show to start for a while uh make sure you can refresh the page and you'll see the link is there I just added that link to the description so if you're watching live make sure to refresh you'll see it now uh but register for the academy and make sure to watch it and thanks for joining us Lloyd um okay so it's live coding happy hour that means we have drinks um I usually prepare some like Spirit or some cocktail or something like that today I'm just drinking a salted caramel espresso um mostly because I've only had two hours of sleep so I'm I need to drink coffee somewhere before I um completely just zone out so I'm drinking coffee today I don't have alcohol uh apologize to the audience who's usually wanting to see what kind of like whiskey or bourbon I'm drinking uh but coffee thank you run off what do you got so I already had like five six cups of coffee today and it's like 12 a.m so I'll just drink water that's all perfect water like I have that at Relic light uh beer it's actually non-alcoholic from the athletic Brewing that they do very good non-alcoholic beer so should be pretty good okay live coating sober hour let's go um uh it happens I feel like that's the reality of uh all of us getting really really busy with work and print off being at midnight so thanks for staying up late for us um that's a good shout out too like we if you pay attention to the content calendar we have shows happening also uh all over the time zones now so some of them are at uh this time some of them are at 1am for me so just pay attention to that calendar and you might have some friendly time zones for you to tune into these shows now and then so that's our dreams that's our introductions let's I'm going to pass it off to pronounce so that we can start diving into the topic of the hour and I'll let pranav uh introduce it all right folks so today's topic is on document intelligence so I think you folks have heard about document intelligence in Tech now also so it is a human assisted parsing system for your documents right you can utilize stock intelligence to parse all document all kind of documents it can be contract notes that are there it can be medical receipts it can be passports that you saw earlier and somebody mentioned in the comments that do we are ready with password so no more password for this time so today I'll tell you about like what's the use case uh for I so it's it's kind of a story last week I went to a nutritionist to get my diet plan so I thought like let's they provided me a receipt or something like to enter what kind if I have any kind of disease or something like that so I got an idea like why not I pass that document and up to it and we just see how we can pass that thing and store in service now so that's the thing let me share my screen now and show you what kind of document I'm talking about so let me know if you're able to see this thing so this is kind of a patient info this is like not the real info this is just like some dummy data that you're there so you can see a lot of check boxes and some fields and all so we have these three uh lists that are that is there so what I'm gonna do is like we're gonna pass the data from these uh these docs so that's the whole idea and let me go ahead and start with the first thing so now what you're seeing is the new experience that just came in Utah so this is our new Doc Intel experience if you have seen the old episode that along in all in shock date in the past so that had a different experience so this is our new experience that we're going to utilize so first thing is like when we have the document intelligence we go to the home page and you can see like there are steps right like set up a use case involve agents monitor over time so personally I haven't played much with the experience so I may make mistakes so I apologize in advance but hey it's a life coding happy hour so we make mistakes here so all right so first thing that we have to do is like if I read it here it's like to set up a use case so you can see like we have a use case uh use case tab here let's go ahead and create our use case all together so I'm going to just create a use case I'm gonna say like okay patient info and I also created a table that is like a patient info tables to store our data so I'm gonna just put it here so you can see this is our table so basically this is the name of our use case and this is a location where the the data will be present so that's there so now you can see right this extraction mode set by default so it's the recommended mode is set by default but there is an another mode that's called as autofill that's there so that comes into picture when we have something like we have some data already collected so that's uh once we have some data that's passed by the AI so then we use the autofill by default servicenow just books like recommended mode here let me just save this thing up several let it load here we have this space patient info that's created now we have to Define some fields that is our first thing so when I talk about field if you see on the screen it can be full name age do you have diabetes high cholesterol height weight whatsoever these are the fields that are there so let's create couple of them and I want to talk about a main highlight of Utah like now you can extract checkbox information from the documents that are there so we have couple of check boxes so let's see if our application will be able to pass these things up so let's go ahead and create like few fields first thing that I'm gonna create full name that's where it will be a text field and I can just say I can keep it empty for now but uh I think I might have name yeah in the table so this field is coming from the target table that we configured during the use case that's there I'm gonna add this thing up hey pronounce before we um uh move forward how did what what did you do to learn how to get all this started like what did you go through just the docs or where did you get a resource form somewhere to be honest like this this application is so easy like I didn't have to go to any docs I just came into dark intelligence and I was like I came to home page and I can just show it again okay so it's just it you just relied on the actual ui's training to get you through it okay that's cool exactly yeah you can see that yeah cool all right how simple it is and let me just go back to our home page and I'm gonna go to the use case I thought you would say di Academy but but yeah I can't be that is something I have to watch to be honest but I'll I think I'm gonna watch it and maybe join next time I wasn't planned by the way I was just legitimately asking like that okay go ahead pronounce sorry all right so we have the full name that's there and now what we're gonna do is like I'm gonna create a group of fields like if you see this thing right uh please provide information about your medical history so I can say the these are group of field like group of check boxes so let's go ahead and create a group so I think you can see like uh it's assigned to a field Group option is there so field Group is something like a collection of similar kind of fields that we can put it there so what I'm gonna do is like I'm gonna create a new one field Group I can say Med history and it is a check box that's there and let me add this thing up so now once I added this thing so you can see it shows up and I can say diabetes like that I'm just gonna say this thing and I'm gonna add a field similarly I can say for cardiovascular disease so I can put it a new field there also so while you have that screen up I yeah I do want to mention that checkbox on the screen that says Define multiple Fields it's pretty convenient when you create multiple fields in a row and you the the so that the pop-up doesn't close and you can create your second field without having to re to to save a click basically yeah I think that's like for mentioning it that's why you have to watch the air currently so it should miss these kind of things all right I'm gonna add this feed so I'll just say two Fields that's all and now you you can see right now I haven't written a single line of code so that's how easy it is anyone can do it so let's see the document task so the next tab that is a document task this is a place where you upload the actual document that you want to parse and store into the table and before I go into this I'm going to go to the integration tab so now once you pass a document so it goes into the like this field value table that's there so I want to go ahead and store this thing in our patient info table so I want to create a record there right so what I'm going to do is like I'm gonna set up the integration first let me go ahead and just click on setup and I can say like process office our process talk I'm really bad at names so apologies on that I mean I'm usually naming them accurate things I think the last time um I was running things I was naming Things based off of what people were chatting about in the chat so it got really it got really weird really fast so an improvement yeah so let's see this thing so now if you see the use case is automatically populated and where do you want to store the document so it's the patient info now you have you can see two options what type of integration you want to set up so we have the process task and we have the extract values so I when I was looking into it so I clicked on this I button and I was like process toss is something like it automatically creates a process extracted as based on specific condition on the target table so I think if there is certain thing that change on the target table we need to use this and the extract value is something like as soon as like the document is processed like uh we the fields will be automatically populated there so for this particular demo I think I'm gonna use extract values because there is nothing like any kind of uh I don't want I don't have any trigger condition or anything and you can see there is an option of create flow here so if I click on save so it might take some time to show up or maybe I need to refresh it so when when we refresh this thing up so there will be a flow that will be coming up so so like the system will automatically create the flow in the back end and you can see that there is a flow that's created and only thing that you have to be sure is like the flow is not active by default like you have to activate the flow and why it's not active because if you want to add any kind of logic or anything to it you can add the logic so right now you can see like it says to try value extracted values and then the set value so I am like not sure about like how set extracted value works but from the name I think it's setting value to the our Target table that's what I feel but yeah this is one of the first times I've seen um flows being created with triggers and actions not from within flow designer automatically like most of the time you have to be in flow designer to start those things up but this looks like Doc Intel is actually establishing a template for you which is actually newer than I've seen yes and this is something that's part of the new Utah release this is the new admin experience that's letting us create initially like we had to as you mentioned earlier we have to go to the floor designers and create it and but now it's there and you don't even have to go to the AES to create it so you can create everything from the experience itself so yeah so our flow is active to be honest I'm not sure like whether uh it will just populate the data we'll see but uh that's how we have set up the flow and let's go to the document task all together so this is the main place where the exact like where the AI model will extract data from our uh like from our documents so let's see so I'm Gonna Save patient one so that's it and I'm gonna add something like and like I have this thing so I have the person named Steve for which I have put the data and as soon as I click add extraction so it will go ahead and start the extraction now when I click on the display name that's there so it opens a pop-up which says open in document intelligence so when I click on it it takes us to a UI where you can see the actual parsing that's there so it's not something that's instant like you upload it and it goes ahead and it will just pass it it will take some time for the air model to look into the document and get the data out of it so till the time it's doing it I think it takes couple of minutes uh based on like what I have been playing around and so till the time the AI model is finding the details from it what I can do is like I can explain you what we have on the right hand side so if you can see on the right hand side we have something like all the fields so you have see the full name is there and we have a group of field like diabetes and cardiovascular disease that's there and on the second tab it's saying to complete like what it means is like we haven't put the information here so that's why it's saying like it's incomplete like we have to complete it and zero completed that's there and there is something like flag and missing so I think flag is something like you can flag a particular field if you want to review it later and the missing one is something if there's a field that the AI is not able to find you can mark it as missing so these are some of the things and you can click on the refresh button on top and you can see the suggestions are already available now so now what I have to do is like you can see see the full name so I just have to type it's Steve and you can see like the AI is able to find steel and you can see like the zero portion on the right hand side that means like why it's zero percent because this is our first uh first document that we are parsing as we add more documents as we train the model the percentage will increase that's there so I'm gonna put C and for the medicine as it's a check box right so as soon as if I click on it you see like the AI model is able to find like these are the three check boxes and this is something a new feature that we added in Utah for dock inteller so now if I click on this thing you can see it got automatically checked and similarly it's able to find this one cardiovascular so both are checked so that's there I haven't added this one it will be in the other docs that I have so you can see now all our three fields are completed and as as soon as I submit it so if we have configured everything we should see some data here in this table and let's see okay so let's now now the actual fun starts what we're gonna we have to look into the flows and everything why the data didn't came so first thing that I can go ahead and see is like it's the actual document task so you can see that this is the actual document task that's there and you can see the status of this task is done let me go ahead and open this document task so I can just go on the application navigator and I can find the latest P1 document task that's there I let me yeah so you can see that the status is done we don't see any kind of error messages or anything that's coming up and we have and we have all the things that are here right so I feel like the document task has all the information but the issue might be with the flow so the flow is active let's see if there are any executions on the floor yeah so there are no execution at all in the flow that means the flow never got triggered so now what we have to do is like we have to look into the flow flow triggers and actions what is happening there so let's see what's happening was that error or like the not no flows Act was that planned or was that like were you actually debugging in a to be honest we are debugging it all right I'm supposed to drink whenever we hit an actual error and stuff like that everyone yeah all right cool just making sure because I was like you you smoothly transitioned in and out of that I was like was that planned I can't tell okay just making sure actually I never thought like uh will be the floor designer will cause issue I was thinking like the parsing might have some kind of a bug that you have to debug but flow yeah so let's look into uh like the document task is there and the flow is having like status changes took done and so stable is this that's there and I think yeah you can see right uh the source table is not getting populated and that's something like it we have already provided the flow table in the use case and it it should have taken it from there so let me see if we can change some of the settings so what I'm going to do is like I'm gonna go into the use case and see if there is any kind of work setting that we need to take so that the source field will get passed there and I don't see any kind of field or extraction mode that is there okay and because Source table should have been passed from here and uh yeah so what I can do is for now I can remove this condition and as soon as the status changes to done it should like retrieve the values from our document that's there and it should set it so let's see in the set value so in the set value we have the extracted values that's there we have the document task that's there and okay so now we don't even have the target record so again I think it will fail so what I'm going to do is for now is I can do like create record so I I want to use the set extracted value because uh it's asking for the Target record and in our document we don't have a Target record to be honest I need to look into this set extracted value what it's doing but for now I can use a create record thing so when you are selling your Integrations as you mentioned there were two different types there's a process task and an extract value if you look at the process task what we're doing in that one is actually creating a document task from your operation table so when you do that you actually have a reference on your document task that link to your patient table and that way when you do the extraction then it knows what you put the values gotcha so I think uh so when when I create when I was just creating the integration instead of like using the exact value I should use the process process task so you need both you need one when it's like an inbound action and the other one is like extraction is like outbound perfect got it so so folks who are watching don't make the mistake that I did go ahead and watch the AI Academy excuse me okay how many times Banner during the Stream that will happen a lot all right so I think uh like let's say a process too really bad in names so this is there and I think we have to create a new flow altogether for it and I don't have to edit it then I hope I didn't save it I don't think I saw you press save okay perfect okay cool so let's see process two is there and the flow is created let me open the flow process too process two final this is the one final final V3 final V3 this is the one click this one that's there and let's see so I think now it will be more of like a created action yeah that's there so the patient information is created but that's what I'm saying like it's saying that patient information created that's the trigger but our patient information is not created as of now so I think it will get started once there is a record that's created in our patient info table but we have to go ahead and create the record first right so let's see I think create record a task that's there and operation record is created create that's what I'm seeing so we have the create task that's there which is like how the display name of process two task definition is there Source record and on okay and then what is it streaming is like it's doing process task and so like I thought of like asking you this thing uh like as I have created a process to this thing so it's saying that we it if the record is created in the patient info then it will trigger right so do I need to create a new flow to create the record first in the patient info table now so you're gonna create your patient you can create a new record in the patient info table and you're going to put the attachment on that record instead of doing it in the document task okay got it perfect oh this is there let me go ahead and activate it then so I see now so we don't even have to go to the admin console to do it it's it's more of like if I create a patient record and I can just attach our this particular patient info it will automatically pass that thing up and store in the relevant Fields yeah the great thing about that is that then you can use a record producer or any of the now platform ways to create record to get to document intelligence which is the beauty of it look is this a good data point for you for like oh this is how someone that's using this for the first time would actually be struggling with it that's actually really good I think we it's yeah it was without reading any documentation we kind of went pretty far that before the before this call we were talking and joking about how uh the show is unpredictable and how far we can get um and the logic points out that if you can get to a certain point without any documentation and without any like pre-prepared research or anything like that and then that shows intuitiveness to your product and this is a good data point for logic hopefully but I was alluding to a past live coding happy hour where I think it was myself Paige Duffy and Andrew Barnes where we were um I dove into a product for the first time without having um with all of my hopes and goals of I I fairly I'm fairly good at this stuff I I think I could figure this also like by the end of the entire hour I made one button and that's all I was able to do um and uh so this is more uh celebrating document intelligence for being more intuitive than um what we've seen in the past so this is great good job to your team look I mean there's a new admin experiences that we see across products they're really helpful to get started without having to to watch too many AI academies right yeah yeah and I think like to be honest I have just played around with dark intelligence for like two or three hours that's all and uh it's so easy like you don't even have to read anything you just play around with it and uh it's it does the job but again yeah if you don't want to make this mistakes you should check out AI Academy so all right so let's get back here so we have both the flows the process docs and the process too so funny names but now what I'm gonna do is like I'm gonna go into our patient that uh this patient for that scare this is our new record and I'm gonna attach the thing like that uh I'm gonna do something under desktop that's there we do we do all of so I have done that now I when I submit it so let's go to the flow designer and see if our process 2 got triggered so it got triggered and I am not sure how much time it will take but based on the document initial document I think it might take another five minutes so knowing before like is there any way to debug it like where is our document and all yeah I was gonna say if you want to get a bit technical I can show you how to see what's happening uh so you'll go to the navigation and you're going to tap the type ml underscore solution dot list solution or Solutions solution yeah and you want to alter them by date like create it or something okay create it this is the one and waiting for training yeah this is the one yeah so every time you create a new document task uh there's a few there are a few Mr solutions that are triggered uh the first one is a passing so it's going to be either a PDF processing or an OCR processing depending on the type of input whether it's an image or a PDF so that's the first uh Mr solution record that you're going to see on that list and once that one is done then you're going to see the one that says the prediction one so the one that says fetching files for training right now it's the one that that's running and that's one okay that's what's going to make the prediction on which fields on your document and which level of confidence we have and and I think like like the the use case that we selected was the recommended one recommendation one it's not autofill so I think like even before the data is created in our page if before it gets filled into our patient info table we have a agent has to manually intervene and like map the field am I right with it um yeah that's that's what the recommendation type is doing um and I think you explained it very well at the beginning you said once you processed a few documents then you can uh switch to autofill and then if the value is confident enough the value the agent doesn't even have to start typing the value is already going to be there that's after like we always say 10 documents right I think is what we always say yeah that's like yeah a general guideline would be to have 10 of the same documents uh it depends it's going to depend like number of fields and quality of the the document and that type of thing quite sure and I think like as soon as the processing is successful there we have we are able to see the process to record that's created and uh you can see like it just opened up and maybe if I type Olive that's coming up for diabetic like it's showing like uh let's see if we have show all check boxes okay so it's showing all check boxes so cardiovascular it's no so we can keep it like this and uh for this one does it understand that there's two check boxes because usually I mean the not checked box usually means no um but it understands that this field has two choices instead of it being a check box so that technically this is a multiple choice then right yeah so I think I played around with this thing and what's happening is like we are having two check boxes that's why yes and no that's there so I was not able to figure out like uh if it's able to parse like it's able to detect if it's an empty one right so maybe like if you have any insight on it you can share um uh yeah I think her Earl was right uh yes and no obviously it's a binary choice you don't really need to have two but but those cases where you have like a group of five or six different check boxes then you can select them as a like it could be an unknown there's an unknown number of boxes that are going to be checked and and that's where uh you you do your selection oh okay yep so like for this particular thing like it's saying no that's there and I think like uh we we have to we might need to have a yes and no field there also like if it's diabetes we have to have like yes or no or that's something we have to create right and I think we have added the yes but yeah it's saying no so I think like the field in here in the fields so I just have one diabetes like I think I need to create another check box like uh of no or maybe like that's so that I can capture the response of no from here and then in our flow itself we can do some parsing and make sure that uh we we are selecting the right thing is it like is it how we should proceed like on this thing yeah one option that could have we could have done here is uh diabetes is actually your group and then the values are yes and then no so you would have two Fields one field yes and one field is no and these two fields are part of a group that you call diabetes got it yeah and I think I showed like when I was creating the group so if you can click on the new field and you can just have a group divinity that's a group that you can create and just put a checkbox list and again you can you click on the Define multiple field and which I learned during this session itself you have the yes you put it as check box and you put like diabetes that's there and you add a field then you say no and you assign it to like sorry you have assign it to this and you just click on ADD so I think we have now two fields that are added yes and no and these are in group so I think I can go from here and use another one on it to Twitter so this is the same document I'm just trying to upload again all of two final yeah and I think this is the one that I just created final final let's see now let's click on it opening dock until I can wait and so again it might take two minutes or so the final final file look is like what are these clowns doing [Laughter] like it's completely having fun like seeing me struggle I mean the the episode the goal of this episode specifically was to call out some of the new Utah features which we did so that's actually that's a good thing that's going on with this but I mean the rest of it is actually on point for us it's just us trying to figure out how to get things to work when we have very little to no knowledge the hard part is um or I mean the not the hard part for everyone else the easy part for us is we have someone like loic on stream with us to actually help us through those kind of struggles but these are the kind of things that um while we wait is a good thing to say like we um we are open to these kind of communications like if you get hung up on something you can either jump into the community um or message one of us um we'll try to provide contact details if we want um it doesn't have to if he doesn't want to at the end but um if you send a question to any of us um on the dev program team um even if we don't necessarily know the answer we can find the person that does know the answer and so communicate with us um don't bash your head against the wall for too long before you start trying to ask for help and we'll try to figure that out stuff for everyone that's what we're here for right exactly and our document is also parsed like I think the I have clicked this button at least 20 times now I was thinking like the alert coming up please stop but that's fine anyway so let's let's see now we have the all of that's there and for the diabetes uh like let's see for this one we will show all that's there so as you see that now we have the group one so for this use case we are saying no that's there and our Let's see we have a med history so I'm gonna say it's a good time to say like okay these are no match found and no match phone I'm just marking it because we already have a group that's created that's there and we have the no that's there and we have something like show all and yeah you have to just say it's there and you know and just say no match or I think it's there but it's checked let me just say how much okay that's there so we have everything that's coming up only one field that's uh that's oh there is one more field to complete that's a yes because I didn't mark it because we already have a no one so let's go ahead and submit it submit anyway all changes are saved and if I go back to the document task the only file to final final will have will have the information that's there let's wait for it to open up so you can see right we have the group that's there which have all the fields and everything you can see for the no one we have the no for the yes it's empty that's there so I think now that's the area we are able to parse it and uh like so I'll just quickly recap what we did so that it's easy for everyone to understand because we did a lot of steps in the middle so first thing is to create a use case that we created that's that's the way you will start once you start the use case then you define your Fields like what Fields you want to parse from the document right in my case it was full name diabetes all these things that were there and then we have something like document task this is a place where you actually upload the document to see the parsing and all and the last tab is the integration so the word integration is doing is like as soon as you create a new record in your target table with the particular attachment the flow will trigger it will do it will run the model and it will go we have to come up back to our document task to uh like map the particular fields and then we have to process it right so that's the whole process of document intelligence and I think those of you who are watching it like you you might have seen like how easy it is to utilize document intelligence and service now so I think that's someone I wanted to show and I'm gonna okay so um as we round out uh Lisa in the chat um wants to say make sure to help us well help the program um product managers create enablement content that is helpful to you if you find in the docs that it's not helpful enough or you want an academy session on a topic that hasn't been covered well tell them and I mean they want to create this content for you so do that check out our uh partner Channel which is community um YouTube Community app on YouTube the servicenow community Channel and that's where all the academies appear um after the fact but also check out any of the registration links because they want you to register at the Academy portal so that they know who's coming who to expect kind of that thing get registration numbers all that stuff so ask join those academies tell them like hey cover this next and they'll they listen to that stuff and they they turn these academies out so pay attention to those um past that as we thank you pranava for diving into document intelligence for us hello thank you for joining us and being the expert and helping us through some of those roadblocks um a few announcements uh there it is knowledge season so check out knowledge.servicenow.com uh it is all in one location again uh huge conference for servicenow um once a year all in Las Vegas in the month of May if you're interested in any of our shows uh make sure to go to the developer.servicenow.com and you'll see all the stuff that's going on there get your own instance to play with for servicenow go to Dublin Utah to find to find our Dev blog which has all of our Utah content including this episode uh if you want to know all the other kinds of things that we do within the deaf program it's devlink.sn links so just go to that and you'll see all our shows um if you haven't yet and you're somehow on this stream without no actually being a subscriber uh make sure to subscribe subscribe and like like the video so that helps a lot too uh uh the hackathon uh website has launched as of yesterday so hackathon is returning uh we didn't have a hackathon last year at knowledge um but this year we have it and we're making it a lot more friendly to our developers and creators and so check that out it's a little bit different than we've ever done before um they uh mistakenly gave me the keys to run the hackathon so I went wild with it and so join us for hackathon at knowledge and uh loic did you have any announcements um I'm gonna guess you're gonna say make sure to check out the academy in the description but if you have any other announcements feel free uh no uh uh everything you say I want to Echo that uh the academies are on YouTube and you can watch them on YouTube that's great if you join live then you can ask your questions and that's even better if you join live at the end you get a little survey it takes two minutes I read every single comments on that so that I can make it better and then I can improve the product so it's not for nothing that we ask for feedback so really I encourage you to to do that if you want um if you want to reach out I'm on the AI intelligence Community Forum every cash every question on document intelligence I'm I'm gonna reply to that so it's the fastest way to to get an answer and then share with the rest of the community awesome Grenada do you have anything you want to announce yeah I'll just like to announce the first inaugural episode on RPA like Asia apga friendly hours will come next week on Friday so make sure you check that out and thank you for like coming and seeing me struggle with document intelligence awesome all right everyone I hope you all have a good weekend um tune in next week remember check the time it's about 12 hours earlier next week for uh different time zone friendliness uh but thanks everybody have a great weekend bye thank you bye final bye final bye final final [Music] foreign [Music]