Full transcript of “This n8n AI Agent will AUTOMATE your Carousels and Slideshows 🍄”. Click a timestamp to open the video at that moment. Need another video? Use the YouTube transcript generator.
0:00 This AI agent automates creating carousels and slideshows and posts it to eight social platforms on autopilot. You can have it run every day or every hour. I'm going to walk through exactly how this workflow automation works. I'm going to show you exactly what you need to do to set it up, which is super easy. And then at the end, I'm going to show you where to get these templates for free for nadn and make.com so you can import them directly into your make or nadn and get going. Real quick, if you don't know me, my name is Sabrina Romanov. I previously sold an AI company for millions of dollars, and now I teach AI for free. You don't have to join my school community to get this workflow automation. And here's how the carousel looks. For example, on Instagram, you have a nice motivational quote here with an image in the background. Here it is
0:46 on Facebook. Here it is on threads. For threads, Twitter and Blue Sky, it'll actually use a long form thread like this, so it's neatly formatted. Here it is on LinkedIn. Here it is on Twitter again as a long form thread. Here it is on Tik Tok. You can see a slideshow. Okay. Here it is on Pinterest as well as a slideshow. And we're going to walk through exactly how this works and how you can customize this to fit your niche or business. So, first I'm going to explain step by step how the workflow works. Then I'm going to show you the only things you need to do to set it up. It's super simple. And then last I'll show you where to download this template for free. So the first note over here, scheduling the trigger. So, this is how often the automation will run. Right now, it's set for every day at 10:00 a.m., but you can customize this run every day or even every hour. The second
1:32 step is our AI agent that writes the motivational quote and prompt. So, open it up here. Okay, I'm going to expand it. So, here's the prompt. Basically, this AI agent brainstorms different quote ideas related to this theme, which you'll want to change later, randomly select an idea, and then outputs a JSON that contains your carousel caption as well as the scene for each carousel slide. What this means is that for each slide in your carousel or slideshow, this is the prompt that's going to be used to generate both the quotes and the image. Make sure you connect your OpenAI account here, right? I'm using 4. And then this structured output parser is just showing how the output will look that will flow into the rest of the automation. All you have to do here initially to customize it is just change
2:19 the theme. So instead of personal motivation, you may want a different theme. And again, this AI agent handles the creation of the text to image prompt as well as the quote itself. If you want more control, like you want to exactly specify the quote that should be used in your carousel, you'll want to enable this second agent down here at the bottom. So, just to clarify, you'll either want to use this top agent here to do everything for you. But if you want more control specifying the exact quote to be used and the exact text to image prompt to be used, then you'll want to enable this second one over here. And let's say you want to switch it. All you have to do is this. This will enable the first one. If you want to enable the second one, all you have to do is this. Okay, I'm going to click test step. So you can see a sample
3:05 output. And here's the JSON output on the right hand side. So this is the caption for example that will be used for Instagram, Tik Tok, etc. And then here's the prompt. Now I'll show you how the second version work. Go ahead and open it up here. Okay. And it's a little different what the output should be. So this text for example is exactly the quote that is going to be used. And then this text to image prompt is exactly the image prompt that will be used to generate that slide. And so this is really handy for example if you want to hook it up to your Google sheet or air table if you have you know a predetermined list of quotes that you know you want to create. Let's go ahead and click test step so you can see what the output looks like. Okay, here it is. So here's the caption and then here's the exact quote that will be used and the exact image prompt. And again, this
3:51 is really handy if you want to override exactly which quote's going to be used or exactly the image prompt to be used. So enable this second AI agent at the bottom if you want that level of control. Then the next step is this setup node over here which has all of your different variables. Go ahead and click this if you're new to NAND. Click this to expand it so you can see. I'll walk through how to set up everything later, but this is the only node you're going to have to set up in order to run the entire automation. So, I made it easy and put all the variables in one place. So, all you have to do is fill out some of these values and then the entire automation will work. The next step in the automation is actually creating the carousel. And for that, we're going to use a platform called blot.com, which I built for myself as a
4:36 solo content creator. It's helped me scale from zero to now approaching 700,000 followers. And it's basically an all-in-one solution to create content, repurpose content, and distribute content at scale. It includes an API that is really, really handy. So, here are the API docs. Just go to helpotato.com, scroll down to API reference, and you'll see the API docs. We're going to use the create video endpoint. And what's really cool is this can create AI videos as well as AI generated slideshows and carousels. So, it's only a single API call that combines AI image generation, video generation, AI voiceovers, scripting, and lots of different image and video models to choose from. How this endpoint works is you can upload your custom
5:22 script or you can use a predefined template. For example, I have template for viral POV videos. We're going to be using this template called quote cards. Here's the relevant documentation for carousels and slideshows. Basically, we pass in the template ID, which is base/slides/coke. We pass in the scene. So, you can think of the scene as each image in your slideshow. Where do you want the captions? Top, middle, or bottom. And then for each scene, this is where there were two variations of the agent. And so either you you can put in a prompt and the AI agent will transform your prompt into a quote as well as a text to image prompt or you can override this and pass in a text to image prompt and text if you want more granular control over exactly what quote will be
6:08 used and exactly what image prompt will be used. So go ahead and sign up for Blot and then what you're going to do is go over to settings, generate an API key. Okay, go down to API access. Keep in mind that API access is a paid feature only. And this helps me reduce spammy content and abuse of service. So don't generate an API key unless you're pretty ready to build this automation. And then after generating the API key, what you're going to want to do is just log in with your social accounts. You can connect multiple social accounts, for example, multiple Instagram accounts, multiple Facebook accounts, and on the starter plan, you can connect up to 20 social accounts. So it's pretty generous. Going back to the automation, so create carousel. All this does is calls the Blotato API with your API key
6:54 and it has all those parameters that I mentioned here. ID, scenes, caption position, script, style, animate, and which text image model you're using. Again, you don't have to change anything here. I'm just explaining how the workflow works. Okay. Once we submit the request to Blotato API to create the carousel/s slideshow, then we have to wait. How long you wait really depends on the image model you're using. When you download this template and import it, it's going to be set to the cheapest, fastest image model. So, that's really helpful for testing, you only need to wait like 1 minutes. Um, but for higher quality images, if you want to use a better image model, you might need to increase the wait time to 3 or 4 minutes. Once the wait time is over, we fetch the completed carousel and slideshow. So, this just fetches it
7:40 with the ID from the Blot API. And then the last step is to publish our completed carousel and slideshow to social media. So when you import this template, only one of these nodes will be activated. And that's because I recommend testing publishing to one social platform at a time. So for example, deactivate Tik Tok and activate Instagram in order to test it. You can go ahead and open this node, but you don't have to change anything here. Everything is already set up and ready to go. So that's the highle flow of how the automation works. First AI writes our carousel and slideshow contents and then AI generates the carousel and slideshow by calling the blot API. Once it's done, we fetch it and then we publish it to eight different social platforms. There are only two things you need to do to get this up and running for your very first time. So if number
8:27 one is obviously connect your open AI account here so that you can use chat GPT. Number two is to fill out the setup node. Once you fill this out, you don't have to touch anything else that comes after this node. It's already set up, ready to go. Everything you need to change is here in this setup step. So, go ahead and open that and then click this on the bottom right to expand it. Okay. And the first thing you're going to get is your Blot API key. So, as I showed you, go to my.bloat.com settings. Grab your API key here. For the template, you don't need to change that. For the script, this will be used as your video title. So, I've just put a placeholder here. Video title visible only in your Blow video dashboard. Feel free to change this. And then the caption and scenes were both generated by our AI agent in the
9:12 previous step. So you don't need to touch that. Caption position. You can change this. The values are top, middle, or bottom. And it is what it sounds like. So where do you want your captions located in the video style, you can change this. I set the default as realistic. Animate all. This is going to be set to false so that the default version just creates static images in your carousels and slideshows. However, some platforms do support like video carousels. For example, Instagram. So, if you set this to true, you can actually animate all of the images into video clips and then push a video carousel. Text image model just refers to the AI model. It's going to be used to generate the images. So, go back to the documentation here and scroll up to text image model. And these are all your
9:57 options. Uh, this is the cheapest one, Flux Chanel. So, when you download the template, this is what's it's going to have. If you want more realistic looking images, use ReCrash V3. Uh Flux 1.1 Pro Ultra is also very good. Just keep in mind that these two models because they are much higher quality will cost a bit more credits than something like Flux Chanel, which only costs one credit. Okay? And then the remaining variables are just your social account ID. So, pretty straightforward to fill out. Go to Blot, go to your settings. For each connected account, just click copy accounts ID. Right. Okay. Go back here for example that was Tik Tok. So then you'll just paste in the accounts ID here. There are certain platforms where you have to specify for example Facebook
10:42 a page ID and for Pinterest a board ID. So if you go to settings again for example Facebook pages this is the page ID you're want to going want to copy. Copy page ID. Okay. Go back here and then again copy and paste your page ID where it says it here. Pinterest board ID is a little more annoying to get. So let me show you how to do that. Head over to your Pinterest accounts and then click rightclick inspect. Go to network. Refresh the page. Actually, if you're not seeing anything here, click one of these. Click payload. And then you're going to see board ID here. Okay. So, copy that over here to your Pinterest board ID. Paste it here. And that's pretty much it. So, obviously, you don't have to fill out the accounts IDs for the platforms you don't want to use. If you have multiple Instagram accounts,
11:29 for example, you can change this to be like Instagram ID 1 and then Instagram ID 2, so you can keep track of multiple accounts. If you are going to do that, just be careful here. You'll want to open like Instagram, for example, and make sure account ID is using Instagram ID 1, Instagram ID 2, etc. So, yeah, that's the only thing you need to do to get everything up and running. Just to recap real quick, just connect your Open AI account here. so that your AI agent can use chat GBT and then just fill out the things in the setup node, particularly your API key and the account IDs that you want to publish to. And that's it. So, I recommend just testing that. Once the entire flow works, that's when I recommend then going back and tweaking the different parameters so that it fits your brand
12:15 and niche. For example, uh swapping the text to image model to a higher quality one. If you want to change the caption position, if you open the AI agent, for example, you may want to change the niche, or you may want to change this prompt altogether. I don't recommend changing this output format because this gives the AI agent an example of what a good output looks like, but you can pretty much change anything here in the instructions. Thankfully, you don't have to build this all from scratch. I've already built it in nadenmake.com and making it available for free. So, in this YouTube video description, open it up and there will be a link to my newsletter. At the bottom of my newsletter will contain the links to the nen templates and make.com template that you can just import into
13:01 your account. And if you don't know how to do that, basically you start a new workflow, click the three buttons up here, and then say import from file, upload the file, and should be good to go. If you enjoyed this video and this type of content, please hit like, hit subscribe, and drop a comment below. would love to hear from you.
