in this video I'll show you how to build this AI automation that will turn your podcast or YouTube videos into hundreds of viral videos completely automated and add a fraction of the cost that you will pay for services like Opus clips and we'll use AI to detect where your face is in the video and also autoc caption it like they do no editing no hassle just content that works for Instagram reals Tik Tok and YouTube shorts all generated for you in this video I'll show you how to build out this automation step by step from scratch so that we can take your podcast episodes just like this video here of me on the school games podcast and how to automatically convert those into social media Clips here you can see we generated four new engaging Clips with captions all automatically and not only will I show you how to build this step by step but I will also walk you through my thought process on how I build these Advanced content automations by the end of this video you'll have a video clip generating machine that does all of the heavy lifting at a fraction of the cost you'll have a working system to grow your personal brand or your clients or you could take this Automation and sell it to other businesses so here's the plan I'm going to go through and demo this system so you can see how it works then I'll show you how to build out the air table database the make scenarios and also how to install the NCA toolkit which is the free tool that gives us an API that allows us to transcribe media and to cut and crop clips and also caption those clips all automatically for nearly free so I'm going to demo the system build out air table and make help you install the NCA toolkit so you can transcribe and edit videos and automations in a cost-effective way and then we'll finish the the entire automation if you're tired of trying to figure out all of this stuff on your own this step-by-step video will save you hundreds of hours let's get into it all right so now let me give you a quick demo of the system I call content clip magic which takes you long form videos and chops them up into shorter Clips with captions automatically everything starts in the video tab you can add a new row here I'm just going to go ahead and copy this previous example this is an interview that I did with the school Stories podcast so you put in a description and then a link to a video for the system to use I'm going to go ahead and copy this into this new row here and then you're just going to want to define the width and height of this particular video and then the width and height of the clips that you want to generate again I'll just go ahead and copy this once you have this information here there is an automation that will automatically trigger it's a very simple automation here we can go ahead and run that this is going to call your NCA toolkit which is going to automatically transcribe the video once this automation completes it's going to send the transcript to error table and then it will populate the transcription and the SRT in these two columns here and we're going to use that transcript and that SRT file to find clips from this longer video and then also we'll use the SRT file for the various timestamps that it found for the transcripts so that we have precise start and end times just like you can see here for each of our Clips so if this is a long video this could take a while so while that's processed I'm going to go ahead and use our example here except I'll jump back to clips and I'll erase all of these so we can start from scratch and then once you have your transcript URL and your SRT URL and we don't have any Clips then this next automation will run this automation here will scan the transcript and will actually build out the clips so we'll go ahead and run this automation here it's downloading the transcript here it's downloading the SRT here we have Cloud processing the transcript to find Clips to split up if you'd like you can also use chat GPT I just happen to find that Claude did a good job at finding Clips this next module here is simply taking a look at all of those clips and it's just building us a Json that's going to make the processing of the clips a lot easier it's going to put everything into a Json that we can iterate over so that then we can drop those new clips into air table so if we take a quick look at what anthropic came back with here and we look at the text response you're going to see we have these clips here in text format and then what this module did here is it simply took that and it structured it into an array of collections of the clip and the description and we can actually see the clip in here and then it goes into an iterator I was experimenting with using AI to actually find that clip in the SRT but I found that actually doing that manually with old school tools was actually a more predictable way so it is good to understand is that there's good times to use Ai and there's also good times when to use something more deterministic like some of these modules here and we'll talk more about that but what these modules here are doing is simply extracting specific segments of the SRT file to pull from when you look at an SRT file you have all of these segment numbers and so what's happening here is that we're searching the SRT file for the transcripts we're matching up what is in the SRT file with what clip we want to pull and then we're grabbing the starting and ending segment numbers to pull a specific clip from the SRT file so here we grab the starting segment the ending segment and then using the beginning and ending segment this parser here actually pulls the entire clip from the SRT file then we insert those new clips into air table so if we jump over here to Clips we're going to see five new clips so here you can see the transcript and then you can see the actual SRT file that is associated with this particular transcript and the reason why we do it this way is because it's better to find the clips using the transcript but it's much easier to find the start and end times of a given clip using the SRT file so then using the SRT file we will gather the start and end times and the duration of that clip and then once we have this information then we can actually go into that long form video and then cut out the clips so now I jumped over to the automation that will actually cut up the clips again we're calling the NCA toolkit if we run this it should run this five times for each of our Clips so now it's processing those clips and when those clips are done they're going to come back here and you can see that they're being populated now then once all of the clips are in it will run the next automation which will crop the video remember that most podcasts are in a horizontal format and we want to pull a specific clip so we've got to crop out the vertical clip we'll go ahead and run this automation now this automation here is looking at the thumbnail of the clip so if we come into the details of the clip we do have a thumbnail right here and in order to properly get the right clip we do need to calculate the X and Y coordinates of where the faces so that's what that module is doing there then this module simply just makes this information more readable in a Json so if we take a look here and we look at the result we have the width of the image and then the X and Y coordinates of where the face is and then we're updating air table with those calculations and then we go ahead and crop and then scale those clips so here you can see one of the cropped Clips is coming back now we have it in that vertical format and it took that from the more horizontal format so this was the clip notice it's not the full podcast but it's still in that horizontal format and then finally once all of these cropped Clips have come back we can run that final option here which is to use the NCA toolkit to caption these videos for almost free it's going to run for however many times we have actual cropped Clips ready to go so here we ran it two times and once it's done captioning those videos it's going to send them back to air table and then we'll be able to see them pop up right here I went ahead and ran this now so that we didn't have to wait for all of these to finish so we're just just going to see two clips pop up so now we've got these final Clips coming back with the captions we can take a look at them and here are the captions so there you go that's how it works let's get into the build and if you're enjoying this video make sure to like And subscribe to the channel it tells me what type of content you want more of all right so now let's actually build this out I'm going to go ahead and provide a template to this aor table database for free there's a link in the description below click on that link and then once you open it up you should be able to come up to the top here and then duplicate the base I'm just going to go through that process so that you can see how it works and just so I don't get confused I'm going to come to the front here I'm going to put YouTube demo I'm going to remove the copy here at the end I'm going to go ahead and duplicate everything click here duplicate base I'm going to open up that base and then I'm also just going to come up here to My Demo and change the color just so I don't get confused all right so now we're in this demo there's a few things that we're going to have to set up along the way there are a few different automations that are already pre-built in this air table database that we'll need to test and turn on I do have a couple of test rows in here that you can use for your automations here's a long form video and then here's also a short one so that we can test generating the transcripts and the srts very quickly when you're transcribing videos they can take a while to finish up so when you're testing things out it's nice just to use a very short file so that these transcripts and srts come back quickly so that you're not wondering if something's wrong and then once you know it works then it's easier to trust trust the process when these longer transcriptions are running now if you want access to all of the blueprint templates that we create in today's video so that you can simply come in and import those Blueprints and have them pop right up just like that with all of the prompts pre-filled and configured so that you don't have to type all of this by watching the video make sure to jump into the no code Architects Community it's a thriving group and you can get access to today's templates and a whole lot more all right so now let's go ahead and build out the first automation which is going to transcribe these larger videos that we put into the system so we can generate the clips and that's this automation here it's very simple we just search Air table for any new records that need to be transcribed and then we send that over to the NCA toolkit for transcription now before we can actually build out this automation we should go ahead and build out the NCA toolkit to build that out I'm going to go to a platform called digital ocean this is a platform where you can launch your own lowcost servers if you don't already have an account go ahead and sign up I already do so I'm going to go ahead and log in now typically I install the NCA toolkit in a platform called Google Cloud here you you can see I'm running two instances of the NCA toolkit in the Google Cloud platform but when we're processing really long videos the Google Cloud platform will time out in those situations and so when we're processing really large media files and automations it's much easier to use digital ocean because it doesn't have that same timeout problem the cost for running one of these servers depends on which CPU and memory you use the server that I'm using right now is $50 a month you can definitely go a lot cheaper it's just going to be a little bit slower so if you're just testing this out remember you can just Just Launch this for one day get it working test it out and then delete it before you pay this amount and just remember you can get a much smaller server as well so if you don't already have a project you're going to want to come up here to new projects I already have one called API so I'm going to go ahead and use that then the first thing we're going to do is come over here to manage and we are going to create a space so what this space does is it allows you to upload files to their cloud storage so that once we process files once we create transcripts or videos then we can actually store them somewhere to be downloaded so I'm to come up here and create a new bucket I'm going to go ahead and use the New York data center I'll create a bucket name YouTube demo I'll use my API project and then I'll go ahead and create this bucket all right so now that we have that space created I'm going to come back to this in just a minute I'm going to jump over here to the app platform I'm going to create a new tab I'm going to come and create an app and again remember what we're doing here is we're creating our own app our own server our own API that uses the NCA tool kit and again you might not be familiar with this but this is an open- source project it's completely free and open to the public and what it allows us to do is it allows us to transcribe videos and caption videos and manipulate video cut them crop them whatever we need to do for free all we need to do is to pay for the server that we hosted on but paying for this server is going to be a lot cheaper than having to have a bunch of different monthly subscriptions to a bunch of different services to get all of this work done and and if you're interested in this project make sure to get involved cuz there's a whole lot more that this can do than we're demonstrating here in this video today so here for resources we're going to come here to container image and we're going to use Docker Hub so go ahead and select Docker Hub and then here in the repository you'll want to paste this I'll leave it in the description below then you can leave this on latest and then for credentials you can just go ahead and click next looks like I had a typo in mind I'm going to go ahead and fix that but if you copy from below I'll have already fixed that for you then go and click next it looks like we do need to actually type latest here I thought it was just there but we have to type latest now I believe we have everything I'll go ahead and click next now here's where you want to be careful this is where you're selecting which type of server that you want to use I've actually never tested this with a 1 gab server it may work but it may not you'll have to just test that if you want to go as cheap as possible I'm going to go ahead and edit this so here you can see the resource size I'm already noticing that this is going to cause an error right now if you see that little Dash here at the end of the name I'm going to go ahead and fix that I'm going to go ahead and save so resource sius I'm going to check this out what do we got first make sure that you only launch one container these are performance settings that you can set up down the line you can have more than one container to make your service a lot faster and there's a bunch of other options as well like having dedicated instances this is actually a very reasonable price if you are generating a lot of videos at scale I'm going to go ahead and just for the demo I'm going to move this up just to a faster server so that the demo goes quicker after you've set this up you'll want to make sure to save it however I have tested this on the 2 gig ram server it was just a lot slower and if you send a lot of requests to it at the same time it really slows down so again I'm just going to use that again make sure your containers are on one for testing I'm going to go ahead and click save you don't need to change any of this then for environment variables go ahead and click there and then we'll go ahead and go into Global we'll go into edit here you want to put in API uncore key and and then just for testing I'm going to go ahead and put test 1 2 3 now make sure you don't put any spaces before or after this and that you use an underscore and not a dash make sure you're copying this exactly or you're just going to run into issues then we have to install three other variables as well so I'm going to go ahead and add those now this is to connect up to our digital ocean space that we created earlier remember as our system generates new files we need a place to upload them so that you can actually download them and use them so now we're going to come over to the settings we need to take some settings from here on this page and copy them into these sections here I'm going to go ahead and type these out first make sure you copy them exactly as I'm typing them no spaces make sure you use underscores everything must be exactly as I type it same with the values make sure there are no spaces before and after and it's exactly as I have it otherwise you're just going to run into trouble and not understand why it doesn't work so here we're going to type S3 endpoint URL come to the next one S3 [Music] access key come to the next one S3 uncore secret key so I'm going to come back to the spaces I'm going to create the access key I'm first going to come up here to the original endpoint I'm going to copy that come back to our environment variables and I'll add that there then I'm going to go ahead and create a new access key all we should really need is read and write access then you can name the key I'll go ahead and put YouTube demo I'm going to create the access key let create it here now this secret key will only be displayed one time so we're going to grab that right now jump back and put that in the secret if you lose that for some reason then you'll need to just generate new keys and then I'm going to come down here and grab the access key ID and drop that right here and that's all we need now I can go ahead and click next if you want to double check anything feel free to do so resources is make sure you pick the right server server size and for the budget again remember you can launch this for one or two days and not pay that full amount but if you're transcribing a lot of videos and generating a lot of videos this is a very cheap monthly price we'll go ahead and go to info go ahead and go to next and for review just double check everything one more time and we'll go ahead and create the resource at this point it's going to actually build out your NCA toolkit given the settings should just take just a few minutes all right so you should see that come up that green box that it is live if you get any other era you're just going to want to walk back through the configuration and make sure you did everything as is and fix anything that might have been done differently now what we should do is just double check to make sure that this is working properly I'm going to jump over to a software called Postman go to post man.com here's what it looks like if you do not have an account and you're not logged in you can go ahead and sign up for free and then you can go ahead and sign in in the description below I've provided a link that will pull up a suite of tests for your NCA toolkit once you click on that link and you're logged in even if you're in your browser you should get this link here you will want to duplicate this into your own workspace just like that and then once you've done that you should be able to come down here to toolkit and then we've got these two tests that we want to run to make sure your toolkit's working properly again you're going to want to click that link and use your copied version in your own account but we can go ahead and click on authenticate now for the base URL here we're going to come back to digital ocean and we're going to come up to this link here rightclick copy link address and we can come back to postman and we can replace that base URL with the URL from digital ocean make sure there are not two slashes here and then you're going to also want to jump over to headers and we're going to replace this value here with your test API key so we type test 1 2 3 and now if we go ahead and click send over on this right side if it's grayed out that means that you do not have this installed in your own account yet you need to duplicate this into your own account but if you have you should be able to send and then that should send the first request over to our new server and you should get a response like this that says authorized now keep in mind this is the same password that we set up in our environment variables in our new server if we come here and we check out the environment variables again remember test 1 23 is where we added this here the API key now we should be able to come up here to this next section here and we can go to test again you're going to want to replace this base URL with our digital ocean URL L again make sure there are not two slashes there like I have there then we're going to jump back to the headers again you're going to replace this as well 1 two 3 and then we can go ahead and test it and you should get a response back like this response code 200 and then we should see a response with a URL and if we click on that we should see a text file if we open that up you have successfully installed the NCA toolkit API great job so now we are good to go now if anything here is not working you're going to need to stop and go back and make sure that you can get through these points here because if this doesn't work then nothing else in the rest of this automation will work all right so now that we have our NCA toolkit server up and running let's go ahead and continue on with our automation I'm going to go ahead and use this test row that you should have as well I'm going to use this video URL in order for this to work if you have anything here in the transcription URL or the SRT URL you do need to remove those and when you remove those and you have this video URL here if you come up here to generate transcript that row should be here so now let's jump up to make and let's go ahead and create a new scenario and then let's go ahead and connect to air table and search for those records that are in this view here so we'll go ahead and do search for records there are better triggers to use other than a search you can use an instant trigger that will trigger directly from Air table it's a bit more complicated to set up and you can change that on your own I'm going to go ahead and show you how to do it with a search this will run at any interval that you provide if you don't already have a connection to air table you'll need to add that now I'm going to go ahead and use one of my existing connections I'm going to use the Bas I'll type YouTube demo and I'm going to use this new base that I created just at the beginning of this video then for the table I'm going to go ahead and select videos and now here for the view I'm going to go ahead and select generate transcript for the output Fields you can go ahead and just select all you can leave formula and limit alone I'm going to go ahead and click save now I should be able to run this I'm going to right click on this module run this module only and now we are seeing the output we have that one row coming from our video we have our video URL so all we need to do is build one other additional module here we're going to use the HTTP module here you're going to want to come down and say make an API key off request then we're going to go ahead and create some credentials for our new NCA toolkit app let's go ahead and add those credentials I'm going to name this YouTube demo NCA toolkit then for the key I'm going to go ahead and type what we created which was test 1 2 3 I'll go ahead and show that so you can see it test 1 2 3 API placement key is in the header and you can go ahead and leave this as is x- api-key go ahead and create and now for the URL I recommend to continue to pull this from our Postman example API calls here I'm going to go into media and then I'm going to come here to transcribe and I'm simply going to take this ending portion of this URL with the slash just like that copy come back to make I'll drop that here in the URL and then I'm going to jump back over to digital lotion I'm going to grab this URL again copy link address and put that here at the beginning and then again double check that I don't have double slashes I do pull that out the method is going to be a post not a git we can leave the headers and the query string alone now for body type we're going to want to go with raw content type we're going to want to go with application Json and then for the request content I'm going to go ahead and copy this in from my previous example you'll want to stop the video and copy this exactly as I have it just like this you'll also want to change this to parse response now I'll go ahead and position this and open it up so it's easy to see again copy everything exactly as you see here for this here you can use a couple of spaces that's all it is 1 2 3 4 it doesn't actually matter you don't have to have it formatted like this in fact sometimes I find when people are entering these in they end up accidentally putting in an invalid character somewhere here so if you end up getting a bad request an error 400 copy everything as I have here exactly with the quotes and the colon and the quotes and everything and the commas and everything but just remove all of the spaces and all of the enters so it's just one big long line there's no spaces at all you know be real careful if you run into that 400 error I'm going to go ahead and leave mine as is but if you are hitting that 400 err it's simply because something here is not exactly right now mine copied over these placeholders so I'm going to remove these to show you how to get those in there so I'm going to come in between these two quotes here and then I'm simply going to copy in the media URL which is this video URL that's coming from our error table database click that you can also drag it just like this and then for the ID again go in between the quotes and we're going to come here and grab this ID you can drag it in make sure it's in between the quotes and now for the web hook URL you can't use that that was my example you're going to come up here to air table you're going to go to automations and we're going to go up to transcript web hook the automation is off and we're going to click on this and over here you're going to see a web hook and we're going to copy that we're going to come back to the Automation and we're going to copy that in right here now we need that because if a transcript takes more than 5 minutes to run this make automation is going to time out so what this is going to do is it's going to allow that transcription to happen in the background and when it's done it's going to send a request into aor table and then that way we can save the transcript back into our system but we do need to go through this test once you have everything in here I'm going to go ahead and click save I'm going to save this automation you can rename it up here now you probably also want to add a filter here a condition so that your total number of bundles and then we can change this comparison here to is greater than zero so what that means is that if we run this search and there's zero here don't run this but if there's more than one go ahead and run this so let's go ahead and run this automation as long as you set up your authentication connection correctly should open this up and see a status code 202 we can go to the data we can see that it's now processing so it's going to process this video here which was a short video so it should take a little bit less time but you do need to come back to your automations and monitor this here so notice here we just had a success what that means is that it finished the video so we need to take a look down here the body open it scroll down a bit you should see a success and if we look at the response we should see a text URL and an SRT URL so this URL contains our transcription and this one contains our SRT file so as long as you're seeing a success here all you have to do is come up here to test automation we're going to run it both of these should be successful as long as they are then you can go ahead and come turn this on and then if you come back to data you'll notice that it's no longer in this view anymore because we have the transcript URL and the SRT URL and here they are so now if this is working for you what I recommend is that you run this on the longer video example that I have here you'll want to remove these two items here now they're going to show up in the view of course they're here now so now when we run this automation again it will now generate the transcript and SRT file for the much larger video which is important for generating Clips it wouldn't be fun to create clips from a 9-second video here but it was good for testing but now we need to generate the full transcript and SRT for this file so once you have it in that condition go ahead and run this again it'll go ahead and run it's going to take a bit of time again remember you do have to have this automation on now or it won't automatically come back in it's going to take 20 30 minutes for that to process you can always come into digital ocean and just take a look at the activity the insights you know if you you'll be able to see when something's running things can take a little bit longer than you might think you can always come in here to the runtime logs as well this is going to show you that things are coming into the endpoint and things are working if you don't see any errors just have a trust that it's that it's transcribing in the background especially if you can see the activity here so once this comes through we'll move on to the next piece all right so now that you have the transcript URL and the SRT of the much larger video we can continue forward what we need to do now is actually scan the transcript and look for really good clips and then add those clips into our database like we have here so to get started if you already have any Clips here associated with this video you'll want to go ahead and delete these here just come up here and delete then we'll come back to videos and if these are here then if we come to this view generate clip scripts then we should have two different stories here that are ready to go at this point you probably want to delete the test row or just remove the transcripts so they don't show up here because creating clips from that test is not a good idea so now we have just one row here come back to make here I'm going to create a new scenario now once again we're just going to start with an air table search except this time we're going to pull from this view here instead of this one going to go ahead and use a connection that I already had YouTube demo here's the content clip magic for this demo the right base now I'm going to go to the table videos and now this time we're looking for the view for Generate clip scripts really we're trying to identify the clips to be more specific all right select all we can leave formula limit that's fine so now I'm going to go ahead and save it if we run this we should now have the output of that one row here and we have the SRT and the transcript so now this is the automation that we're actually building we don't need to build out this section here this is was a test that I was performing to see if AI could do a good job at actually finding the SRT but I found that some standard old tools were actually working better here so we'll go through that pretty easy automation to build the first thing that we need to do is we have the transcripts as a remote file so we need to download those so that we can actually use those so let's come to make let's go ahead and add a new HTTP module we're going to get a file in this case we're just going to grab the transcript URL and drop it in there we can go ahead and continue forward HTTP get a file this time we're going to drag the SRT URL here save let me just make sure did I get the yeah transcript URL make sure you get that and not the video URL so this is going to download the transcript and then it's going to download the SRT file so we can use it in our Automation and next up I'm going to go ahead and add a cloud module here we're going to create a prompt if you wanted to use chat GPT you can go ahead and give that a try instead for the model I'm going to go ahead and use Sonet for Max tokens I'll just put 3,000 and we'll go ahead and add a message for the user we'll go with user for Content We'll add a new piece of content it'll be text and then we have the text here now I'm going to copy and paste this in from my automation if you want to do the same jump into the no code Architects Community but I'll go ahead and explain everything I'll expand it so that you can copy this from here as well so the first thing that I'm doing here and I'll show you how to build this from scratch is I'm converting the data that is downloaded here into a string it's not that important but this downloads it as a binary and so we need to just turn it into a string so it's looking at that transcription as text so you can find this function I believe on the a page here it is two string if you type that it's going to create this little function here and then if you go inside those parentheses you can come back here to your controls and we can drop in the data from this module which is the transcript don't get confused with with these two here this is the SRT this is the transcript so I'm going to put the data in there and then we can take a look at this prompt so we're putting the transcript at the top and then we're saying please Analyze This transcript above and extract compelling segments that are roughly 1 to two minutes in length approximately 100 to 300 words each clip should and then I describe a bunch of different requirements for picking those different clips deliver complete value make sure each of the clips is independently meaningful so you want to pause the video and look through these and copy them just as I have them here topics of Interest some special requirements these are all geared towards producing the best clip possible reiterate critical requirements then I say to create five unique clips and then I give sample output try to be as detailed as I can and there you go that's it so we can go ahead and click save we can go ahead and build out the next module this time I'm going to use chat GPT we're going to create a completion in this case for the model I'll use 40 mini and now what I'm doing in this prompt here is I'm taking the results here which are in text and I'm going to convert them into a Json which makes the rest of the automation easier to process makes it so we can execute things a bit more programmatically text is often hard to work from so we'll go ahead and create a new message and actually one thing I am remembering before I move on is I forgot the system prompt here come back to anthropic real quick scroll down click on advanced settings system prompt you are a video editor task with looking at a video and pulling out shorter clips from a longer video but that those clips still make sense and are enjoyable to watch so we'll go ahead and save that again come back to chat jpt the r user for the text content again I'm going to copy this from my automation copy and then I'll paste it here and then I'll explain what's going on expand it a bit so it's easier to read now at the top here we have the text response from here it broke because I copied and pasted that so I'm going to erase that and then I'm going to drag in this here so we have the response at the top and then the prompt is saying read the text above which contains the clips taken from the longer video transcript return the exact text from each clip in a Json and then this should say with a short autogenerated four to five word description do not change the clip in any way so I'm asking it to give the clip but not to interpret anything and then I also provide an example of that Json here so that it knows how to actually create it then we're going to go ahead and show advanced settings we're going to change the respon resp format to Json object and then we are going to parse that object automatically and we can go ahead and click save and now this is a good time to go ahead and run this because we need the output from this before we can continue on so let's go ahead and run our new automation it's going to find that one row here we should probably also add a filter so it doesn't go on if there's nothing in that view so now what this is doing is it's looking at the transcript and creating those clips and now we have those clips so we can take a look at the result just like we asked for here are the clips that's the format we wanted got five different clips and so now this is taking that text output and creating something more structured that is better for automations so if we take a look at the output now we've got this result here it's a collection inside clips and then we've got the array of those five Clips with the clip itself and then also the description so we're good to go here now we need to add an iterator go to flow control iterator and now we're going to pass in the array that's right here Clips go ahead and click save now again here we're going to add this line here I tried to do it with AI and it would work sometimes but sometimes it would just give me the wrong answer and so what I found was is that using these text parsers actually got me a better more consistent result for something that's more just mechanical we want Precision in terms of where these clips start and end and this was a better solution in this case so here I'm just going to add three go ahead and show more I'm going to add three of these modules real quick match pattern Advance go ahead and pick that one and then I'm going to go ahead and just clone these these real quick so I have them and then I'll go ahead and just add that air table module we're going to update a record so for the text parser here for Global match leave that alone case sensitive no multi-line yes single line no continue no and then for the pattern I'm going to come back and copy this directly just like that I'll go ahead and expand this a bit so you can copy exactly as it is now remember here I'm using a function I've also got a placehold there so I'll go ahead and just retype this so you can see how that would work I'll erase everything here just want to make sure you do everything that I do if you're trying to do it from the video here I'm going to type sub string open parentheses and I'm going to go ahead and add in these other elements even though they're empty just CU it's easier to see so I'm going to add a semicolon semicolon close parenthesis so now I just need to fill in these sections here so this first section here I'm going to grab the clip that is coming from this iterator right here I'm going to go ahead and grab clip now if you are not seeing this sometimes you have to save the Automation and refresh in order for that to show up if you're not seeing it but if you remember we passed in the clips into this iterator so it knows what to pull and then here I'm going to go ahead and just put one and then here I'm going to go ahead and just put 25 and so now what we're doing here is we are searching for a small portion of the string in a given SRT so that we can extract this start number so we have the clip here and it says and what benefit do you feel right so we're searching for a little piece of that inside this big SRT file so that then we can isolate this starting number and then in the next one we're going to do the same thing we're searching for the ending so that we can find the ending right so you want to make sure you copy all of that just as I have it and you want to replicate what I wrote here using this function here notice here that it's gray and it looks a little bit different than normal text because it was able to convert it into into an actual function you can also access these by clicking into these and coming up here and you could just select substring and it would bring in the function that same way going to go ahead and erase that now from the text we're going to pass in the SRT data that we downloaded here so here we're going to type to string open parenthesis you should see it convert to a function then we're going to pass in the data from Step number three so we're going to come back here go to step number three and pass in this data and then we're going to close the parentheses so now everything should be good to go go ahead and click save now we need to do a setup on this one here Global match no case sensitive no multi-line yes no no and then for the pattern I'm going to copy this in as well again I'll expand it so that you can copy it more easily now we have an additional function that we need to build let's remove everything here I'm going to paste it down here just so I can repeat it so we're coming right in here between this question mark and this parenthesis here I'm going to copy this in because it didn't copy correctly so we got going to be right here I'm going to type subring open parenthesis turns to a function semicolon semicolon close parenthesis and now what you can see here is that we have these little calculations that are embedded in there so the first thing I'll do is I'll just put the clip in there because we know that's going to go right here but now we have these two length functions here so I'm going to add those first and foremost so I'm going to type length open parentheses close parenthesis and then you see here we have a minus so I'm going to come over here where's the math I'm going to grab this minus here put the minus and then I'll put 25 then I'm going to come into this section here in this last semicolon with the parenthesis I'm going to type length again open parenthesis close parenthesis and I'm going to come back and drag the clip into that one in between there and then all I need to do here is also drag the clip right there so now we have our new function here so we'll run that and test it and just double check everything as I have it and then for the text we're going to pass in the SRT file again so here we'll go to string open parentheses we'll use the data from Step number three close parenthesis and we can go ahead and save what we did here is we found the starting position found the starting position then we on the ending position and now based off of those we're going to grab the whole chunk so that we have the whole SRT file so let's open this up here Global match no case sensitive no multi-line yes single line these are both no so for the pattern here I'll copy this from my example we'll expand this so we can see everything again now you're going to want to copy everything as I have it two of my placeholders broke so I'm going to erase those and then I'm going to use these placeholders so we're going to put this one here which is this first one so we'll take the value drag it right where it needs to be right after that parentheses then we're going to do the same thing to this one notice here I've got a parentheses that I opened so I'll just rebuild this whole thing here so right here in between that little carrot and the SLB in this case you have to come up here for the parentheses and you have to add them so go to the gear and then add the parenthesis and we're going to come in here come back to the star and I will add in the value from the second piece piece and I'm going to come over here and add in one so what I'm doing here and you can play around with this when you build this but what I was doing here is I was taking the section from 299 to 302 and then I was also adding one so that it would go one extra segment I don't know that I actually need to do that now that I've perfected how I get these values here so you can play around with whether you want to add an extra one or not or just trust exactly what came out but that's all you need right here so now we just need to fill out the text and that is going to come from the SRT file as well so we'll type to string open parentheses and we'll come back over here we'll go to the data we'll be using this data from Step number three again remember data here is transcript this is SRT so here we want to pull from the SRT close parenthesis we'll go ahead and click save let's go ahead and save it and now I'm going to go ahead and just unlink air table real quick just so I can run this from beginning to end and see how everything's running just to make sure it's good save it anyway run the module it's going to give us a warning just run it anyway it's creating the clips clips are done now it's just processing those into a Json so that the automation can iterate over them so now everything process we can take a look so operation one we passed it in the text here is the SRT and then here was the pattern we were looking for and then you can see we have that little section of the script we're looking for and it came back with 301 it found that 30 one is the right place well I have another says and then it says well I have another so it found this and then got us the number and then we can say the same thing in the second one passed in the ending of the sentence and then it found as 316 which is down here 316 and then in this section here it actually extracted the SRT file here starting at 301 and so now what we can do is we can go ahead and put this back and configure the air table module actually I put update but we want to delete so I'm going to delete that add another module air table create record base YouTube demo YouTube content clip magic table clips now this time for clip description we can go ahead and use the description from the iterator and for the transcript itself you can go ahead and use the clip and then for the SRT you're going to want to use the value out of that final text parser now the start and end time will be calculated automatically in order for that to happen we do need to come back to air table we're going to go to our automations and then we're going to turn on this automation here detect start end time and so what happens here is every time we add a new clip into the air table database if that SRT is not empty then we're going to run a little bit of code here to calculate the start and end time using the SRT start and end time for a given segment so if we have this segment here then that code is going to grab this start time right here and then this end time and then that way we know how to pull this specific clip so make sure that is on and then everything else here we can leave as is go ahead and click save and now let's go ahead and run this module again this time actually creating the clips here in our database in the clips table right here so now this automation has finished you can see that we have three Clips in our database we have transcripts and SRT files the start end time and duration so now we can actually make the clip now you will notice that there's only three instead of five remember here in Cloud we did ask it to create us five clips but every once in a while when you're running this you will find that this will actually limit the amount of Clips we actually create based off of the quality of information that we were able to get from these two modules so this module here will only run if this one was successful and this one will only run if this one was successful so if you get fewer rows it's not really an issue you can always just run this again and generate more Clips just by running this here and it will generate five more Clips but this is to ensure the quality of the final clip if there was something a little bit different in one of these these AI models tried to tweak the output of the clip this is what catches that and it prevents us from just producing clips that aren't useful but again if you want more Clips you can run this as many times as you want or generate more Clips up front to generate more clips into air table and so now we can move on to the next spot and actually generate the clips from the longer video using our start end and duration time values all right so let's continue on you'll want to make sure you save this automation then we'll go back to make and create a new scenario from here we're going to go ahead again and we're going to go ahead and search a table records in this case from our a table database we're now in the clips and we need to actually cut the clips so now we have these three Clips or you might have more that we're ready to process now I am noticing one thing that I did forget to do in my previous automation so I'm going to go ahead and help you fix that notice here that these clips are not associated with this video here so if we come here to this video we just created these clips here but there's no actual video clips connected here so we do need to come back to our automation you're going to want to go into the automation we were just in the identify Clips you'll want to reload this Automation and come back here to your aor table record and now we need to come down here to videos we're going to add a item here and for the record ID we're going to go ahead and pass in the record ID from our search so what this is going to do is it's going to connect these clips back up to our video so I'll go ahead and click save I'm going to come back to our database going to go to the clips I'll erase these easy to forget some of these things then I'll come back to that Automation and I'll go ahead and run it again it'll create new clips again remember I had to delete the clips that were already here because this particular view here that creates those clips is looking for videos that do not have any Clips so it's looking for videos that don't have Clips which of course you can change if you'd like to but that's how it's designed for now so I'm going to come back here wait for this to finish wait for it to regenerate those rows that are actually linked to this story here right there and right there you can see that it just added those clips so if I come back here we've got four new clips that came in all now associated with the proper video here so now when we continue on with our automation here that we were building we'll actually have clips that are properly associated with the master video which we need because we need this video here because we still need to go cut the videos out of here and transfer them into this section right here so I'll come back to make again I'll use a connection that I already have existing for the base YouTube demo content clip magic in this case we're going to search the table clips and we're going to look for the view where we need to cut the clips that is this one right here so it's going to process anything that's here and always feel free to investigate the filters here to explain why things are showing up when they do we'll go ahead and select all the fields we can leave formula and limit alone go ahead and click save now if I run this it should bring back those four Clips see right here here bundle one bundle two bundle three and bundle four so we're good to go now the only thing we need to add for this one is a single HTTP call to our NCA toolkit this is going to actually go to the video and based off of the timestamps it's going to cut the video from that Tim stamp and just give us back a video of a shorter length from 1 to three instead of the whole video so I'll go ahead and add the HTTP module right here we're going to make an API key off request test again I can go ahead and use the credentials that I created before YouTube demo NCA toolkit for the URL I'll go ahead and copy that from this test automation you'll need to change the base of this URL to your digital ocean server I'm going to grab this this is mine paste that into the URL so you'll want to grab your url from here and you'll paste that in right here so that your NCA toolkit URL is right here and you don't have double slashes here but then after that it should say slv1 ffmpeg SL compose and then you'll want to change method to post and so what this API call is going to do is it's going to cut up that video right where we said at this start time at this end time with this duration and then it's going to add that clip here so we need a post we don't need any headers or query string for body type we're going to go ahead with raw for content type you're going to need Json application Json and for the request content I'm going to copy this from my Module come down here paste I'll expand this so you can see and copy it from the video you'll want to copy this exactly as I have it here and feel free to use spaces so that it's indented the same way so I have some placeholders in here I'm going to go ahead and remove them so that I can show you how you'll add them yourself I'm also going to remove this web hook because we need to replace it with our web hook we'll do that in just a second feel free to just slowly copy this from the video Just as I have it and then we can come back here for the ID you're going to want to use the ID from Air table you can just drag that in right between the quotes for the file URL going to want to use the video URL right here you can just go ahead and drag that in between these two quotes that you put for the file URL and so here we have the start and the duration we don't actually need the end time we just need the start and the end time so I'm going to come over here grab the start time put it right here then I'll copy in the duration right between the quotes and then I'll start to scroll down I think we only need the web hook I'll double check mine here here's the source ID video URL start then we just need that web hook so if we come back to airt come up to automations we're going to come to cut Clips this time and we're going to click into the trigger here and grab this URL and we will drop that into the web hook URL section here that we talked about so now you got the ID the input URL the start time the duration and the web hook that should be all that we need we'll go ahead and click save I forgot to do this on the last one so you'll want to make sure you do this to each one but come here and add total number of bundles come here just change this to greater than and then put zero that way if there's there's no clips it won't try to process anything I'll go ahead and save it and just for testing I'm going to come in here and I'm going to limit this to one cuz I can only process one right now for the test so I might as well just only run one I'm going to run this automation it's going to run it to the toolkit 202 looks like everything's okay I do realize here that it's giving me the data back as text I'm going to come back in you want to make sure you change this to parse response click save so now let's go back to air table and we're going to want to wait for that process to finish and then when it does you're going to see this popup right here that it was able to clip from the bigger video now again if you're ever wondering you can always come over to your NCA toolkit you can take a look and see if it's running anything do you see any activity does it show that it's actually running the process that you just requested and we can see here that it sent it to our web hook VM sometimes it takes a second to come over so be patient but there you can see it came in I can take a look at the body looks like it was a success so once you see this then you can just come up here to test Automation and you can run that Automation and then everything should be configured and once you do that you should be able to turn it on we should be able to come back to the data and we'll come back to all clips and we'll see we have one of the clips from the video you can click on that open it up and then you'll actually see the clip 40 this one's 46 seconds long so now we can come back to our automation change our airtable search back to 10 because we were just testing one click save and then I'll run this Automation and we're going to see it run those remaining three and then once those are done processing those will come back into our web hook and then we'll have the clips for the rest and then we can continue on with the next Automation and there's that final clip so now we can continue on with the next automation that will actually crop those larger clips into vertical videos so next up we're going to build this Automation and what this automation does is it helps us actually crop out and scale the clip from the long form video so in any given podcast there's usually a guest at some X and Y coordinates so what this first module here is going to do is it's going to figure out using the analyze image module in open AI it's going to figure out where the person is located in that video so that then we can crop it out so first we need to figure that out so we can crop it in the right spot and then we're going to send another request to the NCA toolkit to crop it and also to scale it because this video is 1920 by 1080 so this section here is only 1080 but vertical videos are actually 1920 from top to bottom so not only do we need to extract it but we also need to scale it up so that it it's the proper format on Tik Tok or YouTube and all that stuff we'll jump back to make and we'll create a new scenario and now of course we're going to be using the clips that we've previously generated where we have the clip URL keep in mind in the background it's hidden but if we turn this on here thumbnail URL you're going to see that it also oo generated a thumbnail and this is the image that we actually send to chat jpt Vision so that it's able to isolate the X and Y for the face we also calculate a bunch of other values here just to Ure that we crop in the right spot you can imagine that if you have a clip here and the person's right at the edge here right there if you calculate that spot there might not be enough room over here to have a proper clip so in those scenarios it still has to build the clip here ideally you're not in this situation hopefully you recorded it so you're not way off to the edge but we do need to make some of those calculations and that's what the aot table database is helping us do in the background with some of these formulas feel free to investigate them to see how I went about doing that so let's jump back to our new scenario and the first thing that we need to do is just search our air table View and we'll be using this one here crop clip so we're going to search that view again I'll use a connection that I already have pick the base YouTube demo content clip magic we're going to search the clips and in this case we're going to search the view for crop clip go ahead and select all save then we can go ahead and add our next module we're going to go with open ai go ahead and show more and we are going to analyze images with vision should already have a connection go ahead and update that now I'm going to go ahead and copy over what I have from my prompt going to come into this section here copy and paste and then we'll go over here expand this a bit so you can see analyze the image in great detail and get the height and width of the image then analyze the image in great detail and provide the X and Y coordinates of the center of the face spend time thinking about your answer and then we can map in the image itself how would you like to specify the image we're going to pass it a URL and we'll pass it the thumbnail URL now for the model the only one that I found that worked was C GPT for Turbo I'm sure others work I didn't test each one it was quite interesting using this module in particular every once in a while it would tell me that it was unable to do this kind of work it would say hey I can't actually find the coordinates of a face but most every time it would so it was a little bit bizarre again don't worry about it if you hit errors along the way this automation is here to generate lots of videos so if you hit an error on one clip it doesn't matter you just ignore it and move on because we're going to be generating hundreds of clips from this and single errors should not get in our way and that might be due to the specific model that I picked but in general this doesn't have an issue and works just fine so I'll go ahead and click save I'm going to go ahead and run this just to see how it's performing should grab a few different clips and now it's going to run chat GPT on that we can go ahead and stop it after it runs one here so here you can see it exported all of the details that we were looking for again I'll add a filter here so that it only runs if there are actual bundles greater than greater than zero save then we've got our Chad GPT module now I'm going to add another module here just to analyze the text and the answer answer and to give it to us in a Json it just makes a lot easier to work with I'll come to my existing Automation and we'll copy everything out from here I'll go ahead and copy we're going to create a completion you'll add your connection for the model I'll go with 40 then we'll add in the message roll user text this is where I'll copy and paste what I had before for the result my placeholder broke so I have to add the placeholder from the previous module just drag that in here so basically I'm just saying hey extract the image to dimensions and the XY coordinates from the text above and this text is the analysis of that thumbnail of where I am located in it and I say provide the dimensions in Json format width image width height image height x x value and Y yv value so I'm being very specific on how I want that output so that the rest of my automation has this consistent Json to work with to make it easy so now I'm going to come down here to show advanced settings going to go to Output we're going to go to response format we're going to change that to Json and then we will parse that Json and that should be everything that we need we can go ahead and click save and then I'll go ahead and run that because we do need those values here to build out the rest of the automation so we have some iterations coming back now let's look at the result so you can see here it's just giving this formatted Json that makes it real easy to use these values in the upcoming modules so then next up what we need to do is we just need to update air table with these values by doing that it's going to update some of these values here that we can then use in this module here so by updating it these functions will update and then the output of this module we can use here and it's going to be using the updated values from these functions here notice that these all have errors they won't once we run this update so let's come here we'll add the air table update a record again select your connection base YouTube demo use the content clip magic demo table we're going to update the clips so for ID we're going to use the ID from Step number one then we can scroll down for assessed width we're going to use the width here for the assessed height we're going to use this height right here for the assessed Center x width we're going to use this here and then for the assessed Center y height we're going to use this here so basically we know the size of the image and we also know the X and Y coordinates within that image so now we can build the proper values to actually generate the image looks like I didn't save this module I'll come back here and save it so again that's going to update air table with those values and then when it has those values it can actually calculate these values and then once those are calculated then we can use those calculations here in this next module so I'm going to go ahead and add that next module in this case it's going to be an HTTP module again we're going to use the make an API key off request we'll go ahead and use the credentials we made earlier now for the URL I'm going to copy this over from my working example I'll paste that in here again remember you need to use your digital ocean URL come here grab it copy link address going to replace this this part here with the URL and remember we want this part here to be your digital ocean URL slash not two slashes just one V1 FFM Peg SL compose now this API endpoint is going to automatically come in and grab the video segment and then it's also going to scale it up to the proper Dimensions so again super cool that we can do this in our automations for free using the NCA toolkit so method I'm going to change that to post you can leave the headers and query alone for body type we're going to go with raw for a content type we're going to go with Json application Json and for the request content you're going to want to copy this in directly as I have it going to come down here paste I'm going to change this par response to yes I always do that again there's a few things that we need to change here I'm going to open this up scroll to the top so you can copy this but before I do that I'm just going to remove these placeholders just like that just like that just like that so that you can copy this exactly as I have it here and then you can replace these back in here just need to be super careful but everything should work as long as you get it exactly as I have it here and again if you just want these templates so everything is here then just jump into the no code Architects Community you can download those templates remove this web hook URL so again I'm just going to open this up you want to have this exactly as I have it here filters output with the web hook so now we just need to come back here and replace these values so for the ID we're going to use the ID from Step number one here again you're just going to put that right between the quotes you can drag it or just click it right there now for the file URL we're going to want to use the clip URL so we're going to scale up the clip right so we're going to take that drop that right in here make sure it's right between those quotes now here is the hardest part I'll go back and forth between my example here so that you can see what it looks like when it's done perfectly and then you can copy that into yours the first one we have here is the crop x width and then we'll have the crop y height after that little colon there so we'll come here and again this is at the equal sign let me double check so equals and then we're going to do crop x width so we got crop x width here just going to click that then I'm going to go after the colon then I'm going to add crop y height then I'm going to go in between the next one what do we got next we got clip left x width actually I just realized I made a mistake I was pulling this from module one but we need to pull these from module 4 the update so I'm going to scroll back up here we can take the ID from one but everything else should really be taken from four so I'm going to take that back out can I come up to four I'm going to grab the clip URL from here cuz these hadn't been created yet so a good catch there remove remove so I'm going to go right to where it needs to be crop x width then the next one crop y height then I'll go to the next one and that one is clip left x width clip clip left x width then we need clip top y height clip top y height then we're going to come to the equals and here we're going to add clip x width from videos so that's here clip x width from videos and you want to expand this and grab that inner value here and then you're going to go after that colon and do the next one I believe which is clip y height from videos clip y height from videos again expand it drop this in there and now we're good to go so let's just double confirm that everything is right here crop x width crop y height crop x width crop by height and then remember here we got the equals colon colon colon so then it's clip left clip top clip left clip top and then clip x width and clip y height from videos clip x width and clip y height from video so we got that good after the equals so that should be good then the last thing we need to add is the web hook URL here so now we can just come back to air table going to go to our automations we're going to crop clip we're going to click into this we're going to grab this web hook copy it back here into the between the quotes and once you have this you should be good to go if you run in any errors it's likely because something here is just a little bit off make sure every little thing is here the quotes and the brackets and the colon and all that stuff and again if you ever run into issues where it's saying that there's a bad parameter it's usually because something isn't typed in in right or there's some sort of character here that you can't see that is causing an issue and in that case what I recommend is just removing all of the spaces and enters until it's just one big long line and then you won't have any hidden characters here that might be causing an issue in your automation I'm going to go ahead and click save and then in this case here I'm going to come back and just limit it to one again so that we don't process all of these at the same time just for testing click save let's go ahead and run this got the dimensions figured that out and now we made the request to the NCA toolkit it status 202 that looks good looks like it's processing it so when it's done we should see this pop up here that the test is complete and again remember you can always come back here to see if things are happening you can see that we just made another request then you can always look at these graphs to see if anything's happening the graphs are a little bit delayed so you don't always see things right as they're happening but depending on what server that you're working with that you created when you started up these processes might take a couple minutes to finish successfully all right we can see that that test came back successfully here let's go ahead and take a look at the body we do want to make sure it looks like a success here as well it is possible that you would get an error in which case you'd have to debug what that was so once we get this we can come back to the top we can test the automation we can run the automation that's going to update the air table database with the information we can turn this automation on now anytime we run this again it will run and we can come back to our automation here and we can adjust the limit to 10 again so it'll process 10 at a time go ahead and save it I'll save this here and then we can go ahead and process the rest of the clips while that's happening and then we can jump back to air table jump back to the data and we've got our Clips here now if I scroll over notice it's filling in the data that used to be in error so that it has the coordinates to figure everything out we have our one cropped clip if we open this up it's going to be a vertical video with me positioned in the right spot but again with no captions that's the last thing that we need to do those remaining videos will process and then we'll go ahead to that final clip where we create the cap and again here while we wait you can see that it's calculating the X and Y position of my face for each of these different clips while you're testing if you are using a slower or a smaller server this could take longer when you're processing more videos at a given time once you get this up and running and you've tested it then you'll start to get a feel for how fast things are and how much budget makes sense for you and your specific situation and it's also worth pointing out there are other ways of setting this up that can be even cheaper in this video I'm just covering how to get this set up as quickly as possible now we can see the rest of these cropped Clips coming back again be patient it can take a little bit of time for these to come through especially if you're running four or five operations at the same time the more you run the slower it goes and then they all kind of come back at the same time now while these are finishing up we can go ahead and move on to the final piece which is to caption these videos now any clip that has a cropped clip is now going to show up in this caption view here so now let's go ahead and build out our last automation to finalize everything while these other Clips are coming back looks like we just got another one jump over to make and we'll create a new scenario we'll go with air table and we'll search for records this time again we're going to go with all of the records that are in caption clip got three right now pick the proper connection then the base YouTube demo content clip magic we'll go with the table clips and for the view we'll go with with caption clip go ahead and select all and for the limit just for testing I'm going to go ahead and move this to one right now again when you're done with testing you always want to move that back to to 10 save now we only need to make one HTTP call here so I'm going to go add the module we're going to make an API key off request again you're going to want to use your credentials and then for the URL I'm going to come over to My Demo automation I'll copy this over here into the URL and again you'll want to change your URL to your digital ocean server copy link address put that in here again make sure you don't have two slashes here just one slash I'll expand this so it's easier to see so what we should have is your digital ocean server slv1 SLV video/ caption we'll change the method to a post you can leave these as is for body type we'll go with raw for content type we'll go with Json application Json go ahead and just change the parse response to yes request content I'll copy this in from my working example select all copy come back put that into the request content try to make this as big as I can so that you can see everything in this case it's mostly you're just passing a video URL the settings for the captions like the size and the Bold or what font you want to use and then of course our web hook so let's remove this I'll remove these placeholders just so you can see how you would add them in after you copy and paste everything that you see here in this request so for the video URL we are going to pass in the cropped clip you want to expand that and you're going to want to use the URL here drag that in between the quotes cropped clip URL then you're going to scroll down for the ID you'll want to grab the ID from Step number one I didn't get it in between the quotes so I'll try again you can just put it in between the quotes as well and just double click and then for the web hook URL come back to air table go to the automation come to the caption clip web hook click in to the first step grab that URL and then we can put that right between these two quotes here once you have that you're good to go go ahead and click save I'm going to save it so now this should just run one here and we jump back to our data here we have four now but it's just going to only run one run this automation we can check out the response 202 data looks good so that's going to take a little bit of time to process and then when it's done just like our other automations it's going to hit this web hook we'll see that test down here we can test it turn it on and then we'll be completely done you can see here it just came back with a response from the web hook looking for a success that's good we can come up here test the automation run the automation it's going to run both of these we jump back to our database we're going to see we now have the final clip here with the captions you can see that clip pretty cool then we can come back to our Automation and we can process the rest of those clips looks like I forgot to change the limit here I'll put that back to 10 save run that make sure to turn this automation back on and there comes in those last two clips with captions now if you want these automation templates so that you can simply import the blueprints and have them pop up just like that with the values and prompts preconfigured make sure to jump into the no code Architects Community it's a thriving group you can also get text support directly from me and we have calls almost every single day there's a make and air table beginner course a business course you can get access to the templates you saw today plus a a whole lot more and it's a thriving Community Check out all the wins from the community and all the different issues and wins they've had with the community I hope to see you inside of the community either way I hope you enjoyed this video and I'll see you on the next one