Collaborate on Dummy Invoice PDF for Public Relations with Ease Using airSlate SignNow

Dummy invoice pdf for Public Relations

welcome rpa champions i am the rpa champion  and this live stream we are going to be doing   something pretty cool and also thank you  everybody that has joined the live stream and   let me just fix my audio that has  joined the live stream and that has the   ability to chat but i have disabled the chat  so let me try to enable the chat very quickly i seem not be able to enable the chat but i  will be keeping a track off the comments if   there is an ability to write comments or on my  linkedin messages if you have any questions for   the next live stream i will try to enable the  uh chat but anyways back to this live stream in   this live stream what are we going to be doing  we are going to automate an invoice process   now this is a process that is present in  almost any company that is making money   so uh imagine that you are receiving pdf documents  that are invoices and on this invoices you need to   extract some information like the date the number  how much the number of the invoice how much the   invoice is and information like that about the  invoice now you need to put this information maybe   an excel sheet run some reports transfer it into  another application that is pretty repetitive and   boring activities that can be automated now  that's exactly what we're going to be doing   in today's video we are going to be seeing how we  can automate this process now i will be taking it   pretty slow so you will be able to  follow along if so you choose to so   let's get starting started right away so  the first thing that you will need to do   is go to power automate power automate is a  solution from microsoft it is a tool that is   available with different packages from office now  you'll have to check with your system admin if   you are a corporate account or you can purchase  power automate in the cloud for yourself for 15   for 15 a month or you can try it out for 90  days now once you have created an account with   power automate power automate will allow you to  automate repetitive and boring tasks now microsoft   provides two solutions one is power automate  and the other one is power automate desktop   now the main difference between the two  is that power automate that we will be   using today will allow you to automate processes  that are uh that have apis that have data in   the cloud that are not present on your computer  therefore for example let's say we have a cloud   folder where we are saving some information we  have a cloud email address where we receive things   that is a good example of something that we could  automate with power automate on the other hand   let's imagine that we have an application on  our computer and that application is running   only on our computer and we were automated for  that we would use power automate desktop now   i have a bunch of videos about power automate  desktop so don't forget to check those out and   yeah and learn about our automated desktop  as well all right so uh let's get started   immediately so we're gonna let's recap again  what we're gonna be doing we are going to be   we're going to be learning how to  automate an invoice process to do this   we are going to learn a couple of things now the  first thing that we're going to have to do is   we're going to uh build a new ai model form  processing so what does that mean that means that   we're going to be building a uh somewhere where we  can upload some invoices and those invoices can be   can be processed after that we are going to be  training an ai model to understand the data on our   invoices and to be able to automatically read new  invoices when they pop up wherever we receive them   we are then going to be integrating this into a  bigger flow and fingers crossed that everything   is going to be working in the end we should have  a process that receives invoices puts them in a   folder extracts the data from them and puts  them on a google sheet online automatically   so is this going to be magic i don't think  it's going to be magic it is going to be   uh we're going to be doing it right now all  right so uh to get started we are going to   go to ai builder and we're going to go to  form processing now it's pretty easy as this   i have also downloaded some sample data that i  suggest that you can download as well you can   get the sample data from right here examples  go to sample data to get some sample invoices   if you want to follow along really quickly  right now with me so again we are going to   go to form processing and we are going to  name our model we are going to call this invoice model youtube now there is  also one requirement and that is   uh that is that you should at least have  five or more documents with the same   layout now that is the requirement that will  be required to be able to train our model uh any other best practices uh okay so we are  going to see this also later on it's basically   telling us that we can also create different  collections meaning that if we receive invoices   from different vendors we would group each vendor  in a collection and train it on its own invoices   so let's create this model and the  first thing that we are going to do   is we are going to be adding uh we're  going to be adding different fields so let's add we would like to extract from our  invoice maybe the amount so the amount of the   invoice so the first thing that we have to do is  we have to tell our uh we have tell power automate   what are going to be the different things  that we are going to be extracting from this   uh from this invoice so  let's add also maybe a date let's add another field let's  add maybe invoice number we could try something a little  bit more complicated like   address and we're gonna see in a second  where where we are gonna be using this   uh uh these fields all right so that's pretty  much all of the fields that i would like to uh   to extract for the time being let's  click on next and let's see what is   uh what we have to do next so a collection so  right now it's asking us to create a collection   a collection is basically a folder that is going  to be containing all of the data from a specific   specific source that we want to train so in this  example let's add some data to this collection and let me find so here we go let's add some trained data let's add five invoices and let's upload this five  documents now these are all   different invoices they are from the same  vendor but they are all different they have   different number of lines inside of the  invoice different basically different data   the data is more or less in the same place but  the invoices are different so let's also add another collection and let's add our other vendor that we have  contoso so let's insert the train data again   five invoices is the minimum number of invoices  that you need for the uh to train the model on microsoft all right now we have our two models  so uh we are ready to go to the next step   and therefore we are ready to analyze our  documents and tag inside of the documents where   the fields that we chose before are so remember  before we were choosing where is the uh where   is the date where is the amount where is the yeah  we're adding different fields now these fields now   we are going to tag them inside of the invoice now  what is this going to allow us to do this is going   to allow us to uh when we receive a new invoice  this model that we are going to be building right   now it's going to be trained so that it can always  recognize where where the amount is where the date   is where the different fields from the invoice  are and will be able to extend them from the pdf   document and uh yeah and allow us to use it inside  of our automations put it inside of an excel sheet   or do whatever we would like to do so we gotta  tag our collections so let's let's do this all right some good quick tips so all right we  are looking for the amount so the amount this is going to be the amount now let's look  for the date of the invoice so this is going to   be the date invoice number this is going to be the  invoice number and let's find the address should   we use the address of uh of this corporation  let's use the address of of the provider   so all right we have we have all different kinds  of uh different information now let's tag the   information on another invoice so let's tag it  also on this one so it has already identified the   fields it hasn't identified the address now let  me help it do that let's check the other invoice and as you can see all the data is  changing but the address is remaining   the same so let's tag it again so this is the address all of the other  fields have been tagged already perfect and this is the address again all right so now  that we have tagged all the different fields   we can go back from this collection and we can tag  everything that we have inside of collection 2.   so again here we have to tag the date the  amount the invoice number and the address   of uh whoever is sending us this invoice for the  first document since it cannot recognize anything   we are going to tag this as the amount  we are going to tag this as the date where is the invoice number  right here invoice number   and we'll tag this which is actually it should be it should be the one above i guess but  it doesn't matter let's just use this one   so let's tag this and let's move on to the next  one now this is a little bit maybe repetitive and   boring but this has to be done for almost all the  invoices now in this case it hasn't recognized yet   uh the data so i'm going to have to do it manually  again so this is going to be the invoice number   this is going to be date and this is going to be  the address now we can move on to the next one   and yet again it has not been able to  automatically do this for me so i will be doing it manually the idea behind this is that once  i'm finished uh training five uh five invoices   it will be automatically able to  recognize all of the other invoices   from this from this vendor and extract the  information that i want now let's tag the last let's tag the last uh the last invoice so we have  september date and invoice number and just again   as a reminder this is all pretty straightforward  there is no coding required there is no uh you   don't have to be a developer to follow along  we are what we are doing right now we are   uh training and ai model to recognize invoices so  we are not writing any python code we are not uh   we are not doing anything uh anything except  dragging and dropping and following a certain   a certain logic now that we have done uh trained  not trained our model we have uh let's go back   to our models i think i have called my model  something different that was not my model so this is the model that we were working on so i finished uh finished building or  training all the different models so the   next thing that we have to do is we have to  train our model so we've uploaded 10 images   we've tagged different fields we've identified  them on the different invoices and now what we   have to do is train our model now imagine if we  had maybe another 10 or 20 different vendors that   we wanted to train our model on we could we could  also add it in the different collections we could   have created more collections and added more  more invoice types and trained uh and trained   a model that can recognize any invoice that  we receive now let's say that we have random   invoices that we receive meaning that we cannot  train we cannot train it on so many different   different invoices because we have too many  in that case we could do different things we   could either group our invoices in terms of uh  how many invoices we get from different clients   meaning and trying to group the most numerous  numerous models or we could also train a generic   model so we could train a generic model to  recognize almost any kind of invoice and   also additionally if we don't want to use the ai  builder we could also use a custom python code   that has been trained on invoices and use that  to extract the data as well now anyways while   while while this is training our model is going  to be ready pretty soon so in just a few minutes   the uh the algorithm is going to finish training  this model and it will mean that we have created a   model that has been trained to recognize  two kinds of invoices and extract different   information from these invoices now i don't think  this is available yet in the meantime though we   can start building our process what i mean by  start building our process so let's say that we   have a folder on google drive and we would  like to add this invoices inside of this folder   and extract the information  automatically from this invoices so i will create a folder on google drive   so our model has trained in the  meantime so let's just check it out so here it is we can publish it or we can test  it or we can use it in a flow let's just test it   very quickly so i have trained it on this  model on this on this data let's try using   for example a test data so let's  use this test data and it should be   able to identify the invoice number  the date the address and the amount so   it has identified the amount it has identified the  address the invoice date and the invoice number   now this is an invoice that he has never seen  before it is an invoice with new data i didn't   have to do any uh any formatting of the text or  any parsing of the text meaning i didn't have to   copy all of the text and then look for date and  the date and the invoice number this was all done   very very quickly without any knowledge of coding  uh so yeah this is pretty pretty powerful it does   have its limitations but it is pretty powerful  now let's try also the other type of invoice   that we've trained so let's see if this one  is going to work as well so yeah it should   identify the balance the invoice the address  and the date so it has identified everything   with 99 confidence which is a pretty good a pretty  good confidence score all right so we have tested   our model now this is amazing we already have uh  we have already created a model that is uh that is   that is live meaning that we have already trained  a machine learning model it is already deployed on   microsoft meaning that on datavirus it's already  existing uh and we can already use it in our power   automate and power apps flow all right let's use  it in our power automate and power apps flows so let's create a folder right here this is a  google drive document so i will just create   a new folder i will call this folder invoices so in this in this folder we are going to  put some invoices and we're going to see if we can   if we can extract the  information from these invoices   so i prefer to use this one  train and test let's use this one so we are going to connect this folder  and get this file well actually let's let's imagine we have different files excellent  all right so let's create a new flow what is a   new flow for anybody that is new to power automate  and to automation a new flow is just a new process   or a new automation that is going to allow us  to speed this process up or not have to do it   ourselves so i can create different kinds of flows  i can create them from templates automated instant   cloud flows schedule flows desktop flows and so  forth and so forth many different kinds of flows   for the sake of simplicity and also uh for if you  want to follow along the simplest kind of flow is   we're gonna create a manually triggered flow what  does it mean that means that we're going to come   to this page we're going to click a button and our  robot is going to execute and do different things   so this is our blank canvas now we have to add the  different steps so the first step that we have to   do is we have to get our document from google  drive so let's type in drive and right now what   we are doing is we are adding an action actions  are things that you add inside of your process to   automate the different actions that you do in real  life so right now i am looking for google drive   and i will what i would like  to do is get all of the files so i will list files in folder and let's  get the folder now i have already uh   i have already connected my google account  this but that is pretty easy to do you will   just have to go to my connections and add a new  connection all right so let's look for the folder   that we would like to go in google drive so  we are going to be using power automate and   we're going to be using the invoice folder  so why is the invoice folder empty right now there we go okay so what this is going to allow us to do is  list files inside of the folder and then we   could extract the information from the different  files inside of this folder so let's see what this   let's just test this and see what is the output the output of this process it has been a little bit of time i have not been  using power automate so i might do a little bit   of live debugging with you guys all right so  uh we we ran this so we got we got this file and we got the other file so i imagine that  right now we can go back to edit and we can   add a loop oh no i have accidentally exited outside of my   outside of my canvas for building the processes  so to add a step you would we would go right here   new steps so what do we have to do right now we  have to do a loop so why do we need to do a loop how do you how do we do conditioning   okay so it's applied to each so this is a loop  uh what are we going to do we are going to find the the name and let's see if this is going  to work and in here let's add an action   so what is the action going to be the action is  going to be we used ai builder to create a uh to create a model so a prediction model so  we are going to use ai builder right now   and extract the uh extract information from  our from our invoice now it is lagging a   little bit i'm not really sure why is it  lagging so i will exit i will save my flow i think that this is unnecessary all right let's delete let's delete this step  and actually let's delete this step as well   what we're going to do first is we're  going to check if uh no well actually   can we go back no that was uh that was a  mistake for my side we are not going to   we are not going to i want to upload the file  automatically and not get it from google drive all right so let's extract the files again so copy file list files in a folder let's  do this again so let's see the content of this   power automate and let's choose our invoice  process all right now the next step that we are   going to do is we are going to use drive again and  we are going to get file content using the path so well we could also use make it   be like this and get the path now power  automate has automatically created a loop for us   so we should get the file content using the  path now if we add another action and we add   the ai builder and we use the ai builder predict  features we can choose the model that we have   that we have just created so let's look  for our model that we have just created we forgot to publish our  model so our model is not yet   available let's just select  any model for the time being all right let's go back to our models we'll come back to this later and we just  have to redo it again so let's go back to   our models and let's publish our model once  our model is published we should be able to   see it in the predict ai builder option and  we should be able to utilize it in our flows all right this is taking some time and please after the stream is ready uh after  the stream is done if you have any comments or   suggestions i'm sure there's plenty of suggestions  but if you have any observations or that you would   like to see something new in the follow-up  stream or go into more details about anything   just let me know in the comments below and  i read those uh i read those every single   comment that is posted on my video and i try to  answer to each one of those as well i am very   grateful for all of the comments that you guys  have been uh leading on my uh under my videos   and for the really really constructive feedback  that everybody has been giving me you guys are   really awesome i think that you guys should give  yourself a pat on the back for following through   with automation for being curious for wanting to  learn this is a very important and uh yeah it's   gonna it will allow you to uh upgrade your skills  get promoted get a better job and probably not get   replaced by an automation like the one that we  are building right now so let's we are back in   our model let's publish this model meaning let's  make it available to all of our processes and i   don't know if i've said it already but we can also  use our app or our model that we have just created   in powerapps automatically as well that means  that we could create a app that maybe takes a   picture or scans a document with the camera and we  could we could automatically use it on our chain   model without putting together any kind of  complex flow now let's go back to our flows the flow that we were working on before so what we have here is uh oh well this is  not really what we want to do so list files   in folder so the next step we are going to do  is we are going to use drive again and get the get the content of the file get file  content using the path that's perfectly fine it is being a little bit muggy since this is a   cloud solution i'm not sure if it is my  internet connection or if it is okay here we go so let's delete this again let's save  it so far and let's add a new step   so again we have to we are getting the list  of files inside of this folder and we need to   uh get the file content using the path so uh the  file path is going to be automatically available   for from us from a variables pane from the dynamic  content that automate generates we don't have to   create variables or anything so by using the  list of files and folders here we get access   to all of this so here it's asking for the file  path so i will look for the file path and feed   it in the file it has also created this box which  says apply to each this means that it is going to   do this action it is going to get the file content  using uh using the path for each file that is   inside of this folder now the next step that we  want to do is we want to use ai builder to analyze we want to use ai builder the predict feature  and we want to look for our model so our model   in theory should be available here it is right  here invoice model for youtube so the document   we are also going to add here the file content  this is the output of this action right here so let's run this and see the results that we get so we are we need to add document type let's see if we can get the  document type automatically   id name display path last modified media  type let's see if media type is going to work otherwise i will just add pdf manually  and see if that and see if that works so let's run our flow and see what happens so  the idea behind this flow is that it should   connect to uh it should connect to the cloud  to our google drive it should get finished so it connected so far to the google drive it is  right now running the machine learning part so all   right it seems like it's successful let's see the  output of this so uh all right all right all right   excellent so check it out we  have the invoice number value   we have this is the invoice  from the uh from the file   let's open one of these files and check them  out so 1975. so i think this is the other one did we get carried away  let's see the other invoice one seven eight one seven eight five   one seven eight five so it got it got the other  invoice so it is extracting the invoice number   this is the confidence it is extracting the date  november 18 let's make sure that november 18   and what is the due date 58 is the total that is  viewed so let's check out the total that is viewed   amount address and the total  should be somewhere somewhere here   amount value all right so it has managed to  extract all of the information so i can also i can also use this information now all right now  i have seen it like this and for the other invoice   we will just cycle through it see the other  invoice the other invoice is 1975 so this is the   other invoice right here 1975. this is the  data from it it has been extracted but now   what we want to do is we want to use this in  data maybe in an excel sheet or something like   that so let's go back right here let's create a  new google sheet let's call it invoice and let's   paste the information here invoice  so what do we have we have the amount what else do we have we have the  date we have the invoice number we have also the address and that's pretty much it all right so we  have an excel sheet now i would like to   extract this information and put it in here  all right so let's go back to our process   and just as a reminder this has been like  really simple so far to do i did mess up a   couple of steps right here not i couldn't remember  what to do but uh without it's everything it's   very straightforward and very uh very vocal  meaning that it can be very literal it can   be understood just by reading it so all right  what do we have to do right now we want to add   this information to an excel sheet all right so i  guess we're gonna search for sheets google sheets   and we are uh oops i think i was a little bit  too fast there we are going to search for sheets and we are going to insert a row so that is going to be still inside of our  cycle so for each file is going to insert a   insert arrow all right so let's see where is our  sheet located it's located right here this is it   on which worksheet are we going to be using sheet1  and now automatically we should get the different   columns now because power automate is so  smart and so cool we have the ability to   automatically insert different the different  values so the amount value the date value the invoice number value   and the address so let's look for the address  value here we go this is the invoice number value   i guess it is all right we have we have what it  theoretically is required to insert inside of   our excel sheet so what i'm expecting is to get  two rows here with the data from the two invoices   that we have right here all right let's save this  let's test this and let's see if uh if the magic   happens why do i say if the magic happens because  in less than uh half an hour we will be able to   create an almost full um and almost complete  process that can be used for invoice processing now obviously this process is nowhere near  ready it requires it requires many different   steps like checks validations  fail saves and so forth   but i will show you how those can be  implemented very quickly very easily and   yeah without having to know how to code or having  to know any great any great peculiarities about   any specific problems all right so our process has  failed why has it failed let's do some debugging all right i guess that is why it has failed but we  did we did extract the information so the reason   why it has failed uh or why it's giving us an  error it's because right here we are supposed   to be having just these two invoices now let  me rename this and move this uh invoice youtube but as you can see we have extracted the  information now we're gonna delete this and   make it a little bit more challenging and watch it  as it's doing it live all right so let's also just   move this from the from the invoices folder to  the power automate folder and since we have done   that we will need to update our folder as well  so uh right what we need to do is we're gonna   need to delete this step and we're gonna have  to add it again so let's add this stat again let's navigate to our file the our file is in   power automate and it should be called invoices  youtube there we go this is the worksheet   and let's just add automatically the values this  is so nice and so easy that it is just great   invoice number value and let's  add the address value as well   also keep in mind or check out that there is also  a bunch of different attributes and values that   you could extract that this is extracting so yeah  feel free to experiment and find and extract all   information that you might be requiring or needing  all right now let's also make it a little bit   more uh challenging for this uh for this so  let's add anyways this these invoices as well   here so it's just going to have a little bit more  stuff to process so let's add this invoices also and then let's trigger our process  and let's watch it live as it's   filling out the amount the invoice number and and the address i guess for some reason the address  is uh here well we'll see if   what is going to happen all right so  let's test this let's test this manually amazing run flow and once again uh i wanted to  thank you all for watching this live stream for   learning about this how to create automations uh  automation is becoming more and more present in   your workplace right now is being said that  automation is enhancing employees and the   collaboration between the two is fundamental  however we all know that this is just a   step right now that is going that is uh  just transition towards full end to end   automation but anyways as we can see here  that uh the data from different invoices   is being extracted and is being put inside of  this excel sheet now the next steps for this   could be to take this data send it inside  of an email we could do that really quickly we could send an email for each one of these  invoices uh we could also make it so that   once we receive an email and that email  has an attachment the attachment is   saved automatically inside of the folder that we  have right here the way that our robot function   right now is that we have a folder that is  this right here we would go to our flows   and let's name this in something a little bit  more meaningful so how can we rename this flow no okay so let's go back here not really sure how to edit the name of the  flow right now uh i probably can't see it   somewhere somewhere hidden hidden from plain  view but anyways we have uh we have our flow   right now so what we're gonna do is send  an email for each one of those invoices   and this is just to prove how easy all of this  is and after that i will show you another way   of how you could uh use prebuilt solutions  that are much stronger than this but   yeah let me just show you how easy it  will be to send an email so we will just   search for another connector like gmail or  outlook or whatever we would like to connect to it is being a little bit buggy  so let's just refresh the page let's refresh this page and and let's wait more for this to load so yeah  sometimes it can get a little bit buggy a   little bit stuck that is because it's a cloud  solution i mentioned this i've mentioned this   previously but nevertheless we did build something  pretty amazing and pretty uh yeah something pretty   amazing why do i say that this is something pretty  amazing because uh previously to do something like   this it would be much more complicated and also  a lot more expensive you might have to leverage   different technologies from different vendors  combine them together integrate them and so forth   that's why this is so revolutionary if only it  would load a little bit faster and i promise you   i do have good internet all right so uh back in  our uh back in our process what we want to do is   we wanted to add gmail so let's see if now it is  going to work if for some reason it has decided to   stop working we'll just we'll just do something  else something a little bit more interesting so   all right it has decided not to load maybe we'll  visit it later later on but i wanted to show you   this part right here so let's save this and let's  go to solution what we could do is we could save   our flow crea export it as a solution and uh  and import it into any other into any other uh   into any on onto any other computer or on  to any other user meaning that this process   that we have just created and trained and  built we can not only share it but we can   create new solution outfit and export it  uh export it and share it with whoever   we want so i did want to create also i did also  want to show you uh this document automation so what is document automation so let's imagine  that we would like to uh we would like to   create a process that is just like the one that  we created right now but it is much more robust   what does that mean that means that it consists  of all of the parts of the process so for example   uh the part that is that handles the receiving of  the document the part that handles the extracting   of the document the validation meaning uh  all right i need to extract the amount the   date this and that what if this is not  present what should i do the validation   and then the actual exporting and also the  monitoring meaning uh what did i export   how many times did the flow run what were  the exceptions and so forth so microsoft has   created or is in the process of creating uh these  custom solutions that come already pre-built if   you have all of your data on datavirus and on  power automate solutions or oral dynamics or   microsoft solutions this will be much easier to  set up and it will be a kind of click next next   next process what does this allow you to do  this allows you to set up complete end-to-end   guaranteed and certified processes that  have been already pre-built and that   that are available for you to use so this  is the idea or the future meaning that uh   soon you will have different kinds of  processes that have been already pre-built   and there will be a shop or a place that is  just like this where there is right now only   one solution custom documents and invoices that it  receives coming soon but imagine if you could have   dozens and dozens of different processes here like  start with hr recruiting invoice management p l salaries and so forth hundreds of different  processes that could be automated and that   are relatively standardized and if your data is  on microsoft tools or can be integrated via apis   then this could really speed up your time to  market and your development of your processes all   right ladies and gentlemen that is pretty much it  i apologize for not having the live chat available   in the next live stream we will have the live chat  so that you can ask questions so that you can uh   yeah interact i also want to thank you so much for  watching for uh being a part of the rpa community   and also you are really the best because you want  to learn automation as i said before automation is   taking over the workplace and whoever is not  updated and whoever is not learning it will be   left behind there have been hundreds of jobs that  have been uh that have been automated throughout   time and that have disappeared now those jobs  are mostly of the jobs but anyways enough of   me preaching about this i hope that you have  enjoyed this video and i hope that i will see you   in a next live stream or on the next video thank  you so much for watching and i will see you soon