Transcription
Do you want a shortcut to mastering n8n? It's not a 10-hour course. It's this video.
We're going to skip the fluff and focus only on the 15 core nodes that do 91% of the work. If you want to go from a beginner to building advanced AI automations fast, do not skip this.
After building dozens of complex automations for my AI agency, I realized I used the same 15 nodes over and over. Those 50 nodes fall under five buckets that I reveal as we go, with the last one being the most powerful.
First up, triggers. Triggers start the workflow when something happens or at the time that you set. The most simple trigger is a manual trigger. A typical example for a manual trigger is a lead scraper. You can simply set up this workflow with this settings node here, fill out the amount, keyword, and the city, and then simply start the workflow by clicking here. Let me show you an example. I click here now and this workflow starts now. When I go from this workflow way now, and then go back to this workflow, and the workflow is still running. When I click here on Execution, you will find the running workflow here. So the manual trigger is pretty handy.
The next trigger is the scheduled trigger. It can run on different intervals. For example, in seconds, minutes, hours, days, and so on. And here you can set it up how you exactly want this trigger to run. A typical example for a scheduled trigger is a auto-publishing automation to all social media platforms. This way, you can post every day, every hour, or however you like.
And the next trigger is an app trigger. For example, when your new Telegram message arrives, a new WhatsApp message arrives, or a new email arrives, or a new contact is added to your CRM. A typical example for an app trigger is a chatbot.
When you're inside your n8n and click here Create Workflow, the first step you have to add is the trigger. So when I click here, Add First Step, here on the right menu, I have to select Trigger. Here I can select a manual trigger, an app trigger, a scheduled trigger, and so on.
Next, APIs and HTTP requests. So when your workflow gets triggered, it can make API calls. For example, to send a message. When I click here now Execute Workflow, it will send a message to my email, like you can see here. This email was sent automatically with n8n.
In n8n, you have many APIs already included. For example, when you click here, the plus icon, and select here Action and App, you have access to many APIs that are already integrated directly inside n8n. But what if an API is not integrated inside n8n, what do you do then? For that, you use an HTTP request node. Let me give you an example. For example, the Kai.ai API. This API gives you access to AI video generation APIs, AI image generation, AI music generation, and so on. Once you create an account here, then you can select, for example, Sora 2, and then click here on API, and here you have something that's called Request Example, and here you can select cURL. When I click here now Copy, and go back to n8n, and select our HTTP request node, we have here a button said Import cURL. When you click on it and paste this command here and click Import, this will note, as completely filled out, you only need to replace your API key, and you can already use it for yourself. You simply need to adjust this expression here, so it's based on what you're filling out here.
By the way, each of the integrated APIs here is actually based on a HTTP request node. So behind this node is actually an HTTP request node where some of the details are already hidden. So that way you only have to fill out what is needed for this exact API, and you don't have to worry about the details. A typical automation where an HTTP request is used is, for example, for video generation or when you use an external service like Blotato to post on all social media platforms. So I would say this node is the most important one in n8n. When you master this node, you unlock crazy leverage.
And in case you don't know me yet, I'm Ed. I teach beginners and business owners how to leverage AI and AI automation in my school community. He will take you from a complete beginner to an advanced user that can build AI systems. Like, for example, an AI voice agent receptionist, a fully automated UGC machine, or an auto-publishing system that posts to all social media platforms, and many more.
Bucket number three, storage. So your automation's now pulling data but has no memory. Let's fix that. I'll show you the two most popular ways to store data in n8n. By the way, if you go to my channel and click here on the Posts, you see here many polls. For example, which model powers most of your production tasks? Or, which tool do you mostly use to store data in your n8n workflows? And here you see Google Sheets is number one. And the next very popular option is n8n Data Table.
This Google Sheets node has the following settings filled out. Let's open the document on the side here. Like you can see, the document's empty right now. When I connect it to our manual trigger and activate it, you see how a new row was added here. This way, you can store data in your Google Sheets. And when you add another Google Sheets node, for example, here, and select here Get Rows, and then provide this URL here, and select the sheet here, and then click Execute Step, you see all the items from this table.
Another popular option is the n8n Data Tables. It's a pretty new feature, and it's basically a light version of Google Sheets. On this video on my channel, I explain the pros and cons of each solution. Basically, the advantage of the n8n Data Table is that it's directly integrated and you don't need any API keys. And it's much faster than Google Sheets since it runs on the same instance.
Next, universal data processing. You have now the data, but how do you make decisions with it? I'll show you how to use the If, Switch, and Set node to put logic into your workflows, plus four nodes that are often used.
Let's look first at the Edit Fields node. This node can be used to store data. For example, in this case, I've stored a list of employees here with the name and profession. But you can use this node also as a settings node. For example, like in this workflow, where I used this node to store the API key and the different IDs for the different social media platforms. That way, I don't have to provide them in each of these nodes here, and have a centralized settings node here. So if I share this template, for example, with somebody, they only have to adjust this node.
Another way how this node is also used is to extract data. Let me show you an example. Let's select the get many events note here. For example, if I want now to provide an after day, I could look under calendar slots and look for this variable. But like you can see, you have so many data, it's hard to find the right variable. Or I check the extract data note and see the exact information that I need, for example, date here, and then I can simply drag and drop it here. To achieve that, you need, uh, edit fields note, and here simply provide the variable names and the variables here from the right. But if you like this video so far, make sure to check out my 18 most useful n8n tips after building 100 plus workflows.
Next note is the split out note. Let's say we want to process this employee data one by one. For that, we use the split out note, for example, when I connect it to this note here and double-click on it and click execute step, what we can see, now we have five items as output. So instead of one item as output, we have now five. And now we can use the loop over items note here to loop over each of these items. Let me show you an example of Google Sheets. I can click here, the plus icon, type in Google Sheets and click append row in sheet. Let's take this document here, provide the URL here, select sheet. Then we need here the import data. For that, we go back to canvas, double-click on the loop over items note, and click execute step. Then we go back to canvas and can fill out the Google Sheets note. Here we can simply provide the name and profession. Let's go back to canvas and click save. Now I will execute this workflow manually, and like you can see, the employees are now added to this Google Sheets document.
Next note is the aggregate note. Sometimes you don't want to loop over items. Sometimes you want to combine this data to one item. For example, when I click here, now execute step, we have as output only one item. But why should we do that? For example, let's say our car sales AI agent wants to know the complete list of our inventory. For that, he uses the Google Sheets note first and gets 400 items. But if we would not aggregate them to one item, we would send back this response 400 times, but we don't want to do that. We want to send only one response back with all items at once. Therefore, we use the aggregate note.
Next note, the if note. Let's take this example here again. Let's remove this and connect it again with this loop over items note. I will clean this Google Sheet now, but this time instead of adding all employees, we want only add the employees which have the profession coder, so it's only Alice. For that, I click here, the plus icon and type in if, select the if rows note. Then I go back to canvas and select here this note and click execute step. This way when I select again this note, we see here on the left the input starter fields. Now I drag and drop the profession here and say it should be equal to coder and go back to canvas. Now I click save and execute this workflow manually, and I can see the whole workflow runs now, but it ran only two times because the false branch is not connected to the loo- loop over items. So we connect it here again, click save, and execute the workflow here again. And now it looped over all five items and only the coder was added here.
Similar to the if note is the switch note, but it's much more flexible. A typical example for a switch note is message type. For example, when you receive a WhatsApp message, it could be an image, a text, a voice message, or something else. Since an image needs to be processed different than a text message or a voice message, we have different flows for each one of these, and the switch note checks based on the routing rules which output should be used.
And the last universal processing note that you should know is the code in JavaScript. This note is the most flexible one of all of them. When you double-click on this note, you have here the option ask AI, and here you can tell the AI in natural language what the code should achieve. You can reference input data fields using the dot notation, for example, user.email. This note is used when the other notes cannot achieve what you want to do. So before using this note, try to achieve what you want with the other notes here.
And last but not least, the AI agent note. It's probably most important note in the whole n8n platform. This note is the reason why n8n got so popular in the recent time. When you go to Google Trends and type in n8n and AI agent, you see how both topics are correlated to each other. You can provide this note with a chat model, like OpenAI, Entropic, Google Gemini, and so on. You can give it memory, and you can give it access to tools. Tools like, for example, Airtable, Gmail, Google Calendar, and so on. What makes this note so special is the flexibility, because in this note, you can add here system prompt, and here you explain how it should work and what tools it has and how to use these tools. Then based on any user request you provide, it decides which tools to use. A popular example is an ultimate assistant. For example, this ultimate marketing assistant has access to the following tools. This way he can edit image, create videos, combine videos, reframe videos, create blog posts, and post to all social media platforms.
If you want to know more about AI and AI agents, make sure to join my school community where we go into detail how to build this kind of systems and where you get instructions and templates to build it yourself. I hope you like this video and see you in the next one.